+

CN111221896B - User behavior prediction method and device, electronic equipment and storage medium - Google Patents

User behavior prediction method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN111221896B
CN111221896B CN201811428883.0A CN201811428883A CN111221896B CN 111221896 B CN111221896 B CN 111221896B CN 201811428883 A CN201811428883 A CN 201811428883A CN 111221896 B CN111221896 B CN 111221896B
Authority
CN
China
Prior art keywords
vector
term memory
long short
long
network model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811428883.0A
Other languages
Chinese (zh)
Other versions
CN111221896A (en
Inventor
张立成
陆韬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Original Assignee
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Jingdong Century Trading Co Ltd, Beijing Jingdong Shangke Information Technology Co Ltd filed Critical Beijing Jingdong Century Trading Co Ltd
Priority to CN201811428883.0A priority Critical patent/CN111221896B/en
Publication of CN111221896A publication Critical patent/CN111221896A/en
Application granted granted Critical
Publication of CN111221896B publication Critical patent/CN111221896B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Theoretical Computer Science (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Neurology (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The disclosure provides a user behavior prediction method, and relates to the technical field of big data. The method comprises the steps of obtaining behavior data of a user to be predicted in the last N long periods, converting the behavior data into N feature vectors with sequence, inputting the first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector, sequentially updating initial memory states of the long-short-time memory network model according to the i-1 th long-time memory vector, inputting the i feature vector into the long-short-time memory network model to obtain the i-th long-time memory vector, wherein i is E [2, N-1], updating initial memory states of the long-short-time memory network model according to the N-1 th long-time memory vector, inputting the N feature vector into the long-short-time memory network model to obtain an output vector, and determining behavior prediction results of the user to be predicted according to the output vector. The method and the device can mine the long-term characteristics and the change trend of the user behaviors, and improve the accuracy of the user behavior prediction.

Description

User behavior prediction method and device, electronic equipment and storage medium
Technical Field
The present disclosure relates to the field of big data technologies, and in particular, to a user behavior prediction method, a user behavior prediction device, an electronic device, and a computer readable storage medium.
Background
With the development of big data technology, more and more enterprises are devoted to predicting the behavior and demand of users through big data analysis, thereby providing better services to attract more users.
Most of the existing user behavior prediction methods extract features from historical data of users, and the behavior prediction results of the users are obtained by analyzing the features. The method comprises the steps of extracting characteristics according to experience of data analysis personnel, wherein the effectiveness of the characteristics is difficult to guarantee, more historical data are used as recent data, long-term characteristics of user behaviors and variation trend of behavior habits cannot be reflected, and the defects of the two characteristics result in lower accuracy of prediction of the user behaviors and influence on actual application effects.
It should be noted that the information disclosed in the above background section is only for enhancing understanding of the background of the present disclosure and thus may include information that does not constitute prior art known to those of ordinary skill in the art.
Disclosure of Invention
The disclosure provides a user behavior prediction method, a user behavior prediction device, electronic equipment and a computer readable storage medium, so as to overcome the problem of low accuracy of the existing user behavior prediction method at least to a certain extent.
Other features and advantages of the present disclosure will be apparent from the following detailed description, or may be learned in part by the practice of the disclosure.
According to one aspect of the disclosure, a user behavior prediction method is provided, which comprises the steps of obtaining behavior data of a user to be predicted in N latest long periods, converting the behavior data into N feature vectors with sequence, inputting a first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector, sequentially updating an initial memory state of the long-short-time memory network model according to an i-1 th long-short-time memory vector, inputting the i feature vector into the long-short-time memory network model to obtain an i-th long-time memory vector, wherein i is [2, N-1] and updating the initial memory state of the long-short-time memory network model according to the N-1 th long-time memory vector, inputting the N feature vector into the long-short-time memory network model to obtain an output vector, and determining a behavior prediction result of the user to be predicted according to the output vector.
In an exemplary embodiment of the disclosure, the obtaining behavior data of the user to be predicted in the last N long periods and converting the behavior data into N feature vectors with a sequence includes sequentially counting O behavior data of the user to be predicted in the last N long periods according to short periods, each long period including M short periods, and converting the behavior data into N M x O dimensional feature vectors with a sequence.
In one exemplary embodiment of the disclosure, the long-short memory network model at least comprises an input layer, a first full-connection layer, a long-short memory layer and an output layer, wherein the input layer comprises M-O neurons and is used for inputting the M-O dimension feature vector, the first full-connection layer comprises P neurons and is used for converting the feature vector into a first intermediate vector of P dimension, the long-short memory layer comprises Q neurons and is used for converting the first intermediate vector into a long-short memory vector of Q dimension according to the initial memory state, any two neurons in the Q neurons are connected, the second full-connection layer comprises R neurons and is used for converting the long-short memory vector into a second intermediate vector of R dimension, and the output layer comprises S neurons and is used for converting the second intermediate vector into an output vector of S dimension.
In an exemplary embodiment of the present disclosure, the long-short-time memory layer is a forward long-short-time memory layer, the long-short-time memory vector is a forward long-short-time memory vector, and the initial memory state is a forward initial memory state; the long-short-time memory network model further comprises a backward long-short-time memory layer which comprises T nerve cells and is used for converting the forward long-short-time memory vector into a T-dimensional backward long-short-time memory vector according to a backward initial memory state, any two nerve cells in the T nerve cells are connected, the second full-connection layer is used for converting the backward long-short-time memory vector into a second intermediate vector, the step of inputting the first characteristic vector into the long-short-time memory network model to obtain the first long-short-time memory vector comprises the steps of inputting the first characteristic vector into a long-short-time memory network model to obtain a first forward long-time memory vector, the step of sequentially inputting the i characteristic vector into the long-short-time memory network model according to the i-1 long-time memory vector, and the step of sequentially inputting the i characteristic vector into the long-short-time memory network model to obtain the i long-short-time memory vector according to the i-1 forward long-time memory vector and the i-1 backward long-time memory network model, and the step of sequentially inputting the i characteristic vector into the i-1 forward long-short-time memory network model and the i forward long-time memory network model.
In one exemplary embodiment of the present disclosure, s=m+1.
In an exemplary embodiment of the present disclosure, the first fully-connected layer and the second fully-connected layer are ReLu (RECTIFIED LINEAR Unit, linear correction Unit) active layers, and the output layer is a Softmax (normalized exponential function) active layer.
In an exemplary embodiment of the disclosure, inputting the first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector includes inputting the first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector and the first output vector;
The method comprises the steps of sequentially updating the initial memory state of the long-short-time memory network model according to the i-1 long-short-time memory vector, inputting the i characteristic vector into the long-short-time memory network model, and obtaining the i long-short-time memory vector, wherein the sequentially updating the initial memory state of the long-short-time memory network model according to the i-1 long-time memory vector, inputting the i characteristic vector into the long-short-time memory network model, and obtaining the i long-short-time memory vector and the i output vector, and i is E [2, N-1];
The method further comprises the steps of sequentially determining an ith real vector H i according to the ith feature vector, matching the ith real vector H i with an ith-1 output vector G i-1, and judging that the current user behavior prediction is invalid if the following condition (1) is met:
And if the condition (1) is not satisfied, executing the step of determining the behavior prediction result of the user to be predicted according to the output vector, wherein the output vector is an Nth output vector.
According to one aspect of the disclosure, a user behavior prediction device is provided, which comprises a data acquisition module, a model analysis module, a result output module and a prediction module, wherein the data acquisition module is used for acquiring behavior data of a user to be predicted in N long periods and converting the behavior data into N feature vectors with a sequence, the model analysis module is used for inputting a first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector, sequentially updating an initial memory state of the long-short-time memory network model according to an i-1 long-time memory vector, inputting the i feature vector into the long-short-time memory network model to obtain an i long-short-time memory vector, updating an initial memory state of the long-short-time memory network model according to an N-1 long-time memory vector, and inputting the N feature vector into the long-short-time memory network model to obtain an output vector, and the result output module is used for determining a behavior prediction result of the user to be predicted according to the output vector.
According to one aspect of the disclosure, there is provided an electronic device comprising a processor and a memory for storing executable instructions of the processor, wherein the processor is configured to perform the method of any one of the above via execution of the executable instructions.
According to one aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method of any one of the above.
Exemplary embodiments of the present disclosure have the following advantageous effects:
And acquiring behavior data of the user to be predicted in the last N long periods, converting the behavior data into feature vectors, sequentially processing the feature vectors by using a long-short-time memory network model, synchronously updating the initial memory state of the long-short-time memory network model, finally obtaining an output vector corresponding to the N-th feature vector, and determining a behavior prediction result of the user to be predicted according to the output vector. On one hand, the long-term characteristics and the change trend of the user behavior can be mined by collecting N long-period behavior data of the user and sequentially analyzing the N long-period behavior data according to time sequence through a long-time memory network model, so that the user behavior can be predicted more accurately, on the other hand, the long-time memory network model is a special neural network model, and the data of the feature vector can be extracted from the machine dimension, so that the limitation of manually extracting the features is avoided, and the accuracy of the user behavior prediction is improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure. It will be apparent to those of ordinary skill in the art that the drawings in the following description are merely examples of the disclosure and that other drawings may be derived from them without undue effort.
Fig. 1 schematically shows a flowchart of a user behavior prediction method in the present exemplary embodiment;
fig. 2 schematically shows a schematic diagram of a long and short memory network model in the present exemplary embodiment;
fig. 3 schematically illustrates a schematic diagram of a long and short time memory layer in the present exemplary embodiment;
Fig. 4 schematically shows a schematic diagram of another long-short-time memory network model in the present exemplary embodiment;
Fig. 5 schematically shows a flowchart of another user behavior prediction method in the present exemplary embodiment;
fig. 6 schematically shows a block diagram of a configuration of a user behavior prediction apparatus in the present exemplary embodiment;
Fig. 7 schematically shows an electronic device for implementing the above method in the present exemplary embodiment;
fig. 8 schematically shows a computer-readable storage medium for implementing the above-described method in the present exemplary embodiment.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. However, the exemplary embodiments may be embodied in many forms and should not be construed as limited to the examples set forth herein, but rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Exemplary embodiments of the present disclosure first provide a user behavior prediction method. The user behavior prediction refers to predicting whether a user makes or does not make a certain behavior in a certain time in the future, or making the degree of a certain behavior, for example, predicting whether the user gets a coupon and the number of times of getting in the future in one week, or predicting whether the user logs in a client program and the time of logging in three days in the future, and the like.
Referring to fig. 1, the user behavior prediction method may include steps S110 to S150:
Step S110, obtaining behavior data of a user to be predicted in the last N long periods, and converting the behavior data into N feature vectors with sequence.
In the present exemplary embodiment, the long period may be a time period for user behavior prediction, for example, predicting user behavior for one week in the future, and the long period may be one week, and behavior data of the user to be predicted for the last N weeks may be counted in step S110. The behavior data of each long period may be converted into a feature vector, for example, various behavior data in each long period may be counted and arranged in a certain order as a feature vector, or behavior data in each long period may be arranged in a time order as a feature vector, or the like. The behavior data of each long period should be converted into feature vectors with the same form and dimension according to the same rule, and according to the time sequence of N long periods, the N feature vectors can be numbered, the feature vector corresponding to the earliest long period is the first feature vector, and the feature vector corresponding to the latest long period is the Nth feature vector. The larger the numerical value of N, the more abundant the obtained behavior data of the user to be predicted is, which is beneficial to the accurate prediction of the follow-up, but the difficulty of the behavior data acquisition and the follow-up operation amount are also needed to be considered, and the numerical value of N can be set by integrating the factors according to the actual situation.
Step S120, inputting the first feature vector into a long-short-time memory network model to obtain a first long-short-time memory vector.
The long-short time memory network model refers to a neural network model which adopts a long-short time memory layer as a circulating middle layer. In the present exemplary embodiment, a trained long-short term memory network model is utilized to conduct user behavior prediction. The method comprises the steps of inputting the characteristic vector into a long-short time memory network model to obtain the value of each neuron and the output vector, and arranging the values of the long-short time memory layer according to the sequence of the neurons to obtain the long-short time memory vector, namely a middle layer vector of the model. The first feature vector corresponds to a first long-short-time memory vector, the second feature vector corresponds to a second long-short-time memory vector, and so on.
And step S130, sequentially updating the initial memory state of the long-short-time memory network model according to the i-1 long-short-time memory vector, and inputting the i characteristic vector into the long-short-time memory network model to obtain the i long-short-time memory vector, wherein i is [2, N-1].
In this exemplary embodiment, the long-short memory layer of the long-short memory network model has an initial memory state, which can be regarded as a hidden neuron. The first neuron of the long-short time memory layer is determined by the initial memory state and the last middle layer. Therefore, when the i-th feature vector is processed by using the long-short memory network model, it is necessary to update the initial memory state in addition to the i-th feature vector. Specifically, the first feature vector may be considered to have an initial memory state of 0 or may be set to another value by a person, and then the initial memory state of the current long-short-term memory network model may be determined from the previous long-short-term memory vector, so that the long-short-term memory network model is used to process the feature vectors sequentially.
And step S140, updating the initial memory state of the long-short-time memory network model according to the N-1 long-short-time memory vector, and inputting the N characteristic vector into the long-short-time memory network model to obtain an output vector.
And sequentially processing each feature vector by using the long-short memory network model until the last feature vector is the Nth feature vector. When the nth feature vector is processed, the method for determining the initial memory state is the same as step S130. After updating the initial memory state, the nth feature vector is input into the long-short-time memory network model, and the nth long-time memory vector and the output vector can be obtained.
And step S150, determining a behavior prediction result of the user to be predicted according to the output vector.
The dimension of the output vector is usually of practical significance, and represents the classification result of behavior prediction, for example, whether the predicted user gets coupons in the future week, the output vector may be a two-dimensional vector, the values of the two dimensions may represent the probability that the predicted user gets coupons and does not get coupons, or the predicted user gets coupons in the future week, the output vector may be a b+1-dimensional vector, the value of the first dimension represents the probability that the predicted user gets one coupon, the value of the second dimension represents the probability that the predicted user gets two coupons.
In the present exemplary embodiment, the output vector represents the behavior prediction classification of the n+1th long period (i.e., one long period in the future), so that the output vector may be converted into a behavior prediction result, specifically, the dimension with the largest value in the output vector may be regarded as the behavior prediction result, for example, in the example of predicting the number of coupons received by the user in the future in one week, if the value of the C-th dimension of the output vector is the largest, the prediction result may be the user to receive C coupons in the future in one week, and in addition, a certain threshold may be set, and when the largest value in the output vector reaches the threshold, the model prediction may be considered to be successful, and the behavior prediction result may be outputted, otherwise, the model prediction may be considered to fail, and more behavior data may be required. For example, when the predicted user gets a coupon in the future week, the obtained output vector is [0.6,0.4], which indicates that the predicted probability that the user gets a coupon in the future week is 60%, the probability that the predicted user gets a coupon in the future week is 40%, and if the threshold is set to be 70%, the predicted probability that the predicted result gets a coupon in the future week is relatively high, but is less than 70%, the reliability of the predicted result is insufficient, and the result of the prediction failure can be output. The process may then return to step S110, increase the value of N, or increase the kind of behavior data, etc., and repeat steps S120 to S150.
Based on the above description, in the present exemplary embodiment, the behavior data of the user to be predicted in the last N long periods is obtained and converted into the feature vector, and then the feature vector is processed by sequentially using the long-short memory network model, and the initial memory state of the long-short memory network model is updated synchronously, so as to finally obtain the output vector corresponding to the nth feature vector, and the behavior prediction result of the user to be predicted is determined according to the output vector. On one hand, the long-term characteristics and the change trend of the user behavior can be mined by collecting N long-period behavior data of the user and sequentially analyzing the N long-period behavior data according to time sequence through a long-time memory network model, so that the user behavior can be predicted more accurately, on the other hand, the long-time memory network model is a special neural network model, and the data of the feature vector can be extracted from the machine dimension, so that the limitation of manually extracting the features is avoided, and the accuracy of the user behavior prediction is improved.
In an exemplary embodiment, step S110 may be implemented by:
And sequentially counting O behavior data of the user to be predicted in N long periods according to the short periods, wherein each long period comprises M short periods.
The behavior data are converted into N M-O-dimensional feature vectors with sequences.
Wherein, the short period refers to a basic data statistics period, and corresponds to a long period, for example, the long period may be one week, the short period may be one day, and each long period includes 7 short periods. In the present exemplary embodiment, the long period is an integer multiple of the short period.
When the behavior data are counted, various related behavior data can be counted according to the predicted behaviors required. For example, in a scenario where a predicted user gets a coupon, the following 5 kinds of behavior data may be counted, in a short period of one day, whether the user signs in to the coupon getting platform every day may be represented by 0/1, the number of times the user signs in to the platform every day, whether the user gets a coupon every day may be represented by 0/1, the number of times the user gets a coupon every day, and the total coupon value of the coupon the user gets every day. Taking one week as a long period, 5 kinds of data per day are counted to obtain 35 kinds of data, so that the behavior data of each long period can be converted into a 35-dimensional feature vector. And counting a plurality of long-period behavior data to obtain a plurality of 35-dimensional feature vectors, wherein the feature vectors have a sequence corresponding to the long period.
According to practical application, M and O may be any positive integers, and the disclosure is not particularly limited.
In an exemplary embodiment, referring to FIG. 2, the long-short term memory network model may include at least the following neural network layers:
an input layer including M x O neurons for inputting M x O dimensional feature vectors;
A first fully connected layer comprising P neurons for converting the feature vector into a first intermediate vector of P dimensions;
The long-time and short-time memory layer comprises Q neurons, and is used for converting the first intermediate vector into a long-time and short-time memory vector of Q dimension according to the initial memory state, and any two neurons in the Q neurons are connected;
A second fully connected layer comprising R neurons for converting long-short term memory vectors into second intermediate vectors of R dimensions;
An output layer comprising S neurons for converting the second intermediate vector into an S-dimensional output vector.
The feature vectors input by the input layer represent behavior data of M short periods of a user, have certain time sequence features, can be fully extracted at the first full-connection layer, can mine time sequence association in the long-short-time memory layer, are sorted through the second full-connection layer, and are finally converted into output vectors through the output layer.
In this exemplary embodiment, referring to fig. 3, any two neurons in the Q neurons of the long-short time memory layer are connected, and a certain weight is set to indicate that a state can be transferred between any two or more neurons. In addition, a state that can be transferred only in one direction in the long-short memory layer may be set, for example, Q neurons are arranged in order of C 1、C2…CQ, and when i > j, there is a connection weight W (C i->Cj) =0 from C i to C j for any i, j e [1, Q ]. In particular, it is also possible to set the state transferred by the last neuron only, i.e. if and only if i=j-1, there is a connection weight W (C i->Cj) noteq0 of C i to C j. The present disclosure is not particularly limited thereto.
In an exemplary embodiment, the number of neurons in the long and short time memory layer may be on the order of 1000, e.g., Q may be 512, 1024, 1536, etc.
As shown in fig. 2, when each feature vector is processed by using the long-short-time memory network model, the initial memory state in the next round of model processing can be calculated by using the long-short-time memory vector obtained by the long-short-time memory layer, and the next round of model processing is participated. Therefore, the time sequence characteristics of the behavior data of the last long period are transferred to the processing process of the behavior data of the next long period, and the time sequence characteristics of the behavior data of the long period are mined.
In the prediction of the user behavior, N rounds of prediction are performed by using the long-short-term memory network model, where the long-short-term memory network model used in each round of prediction is connected to the long-short-term memory network model of the previous round through the long-short-term memory layer, and the connection weight can be regarded as a cyclic connection weight of the long-short-term memory layer (i.e., a weight for calculating the initial memory state), and for any round of prediction, the weight is a parameter in the model and should be the same.
For the long and short term memory network model described above, a large amount of historical user data may be utilized for training. For example, behavior data of a plurality of users in N+1 historical long periods is acquired, the behavior data of the earlier N long periods are converted into sample feature vectors, the behavior data of the 2 nd to N+1 long periods are converted into behavior labels, and accordingly the corresponding relation between the two feature vectors and the behavior labels is generated, wherein one feature vector corresponds to the behavior label of the next long period, and the other feature vector corresponds to the behavior label of the N+1 long period. The internal connection weight of the long-short-time memory network model and the cyclic connection weight of the long-short-time memory layer (namely, the weight of the initial memory state is calculated) can be trained through the corresponding relation between the two feature vectors and the behavior labels.
Further, the long-short time memory layer is a forward long-short time memory layer, the long-short time memory vector is a forward long-short time memory vector, and the initial memory state is a forward initial memory state, and referring to fig. 4, the long-short time memory network model may further include a backward long-short time memory layer including T neurons for converting the forward long-short time memory vector into a T-dimensional backward long-short time memory vector according to the backward initial memory state, where any two neurons of the T neurons are connected;
accordingly, a second fully connected layer may be used to convert the backward long and short term memory vector into a second intermediate vector.
Step S120 may include inputting the first feature vector into a long-short-time memory network model to obtain a first forward long-short-time memory vector and a first backward long-time memory vector.
Step S130 may include sequentially updating the forward initial memory state and the backward initial memory state of the long-short-time memory network model according to the i-1 forward long-short-time memory vector and the i-1 backward long-time memory vector, respectively, and inputting the i-th feature vector into the long-short-time memory network model to obtain the i-th forward long-short-time memory vector and the i-th backward long-time memory vector.
By adding a long-short time memory layer, the vectors in the first full-connection layer can be fully subjected to time sequence feature mining, and the accuracy of user behavior prediction is further improved. It should be understood that, according to actual needs, a third or more long short time memory layers may be further provided, and other intermediate layers may be interposed between the long short time memory layers, or a third full connection layer, a fourth full connection layer, etc. may be added, which is not particularly limited in this disclosure.
It should be noted that P, Q, R, S, T is the number of neurons in different layers in the long-short-term memory network model, and the number of neurons is determined by the parameter setting condition of the developer and the training condition of the model, and is specifically related to a plurality of factors such as an application scenario, the complexity of user behavior data, the number of sample data, training parameters, and the like. Therefore, P, Q, R, S, T may be any positive integer, which is not particularly limited by the present disclosure.
In an exemplary embodiment, the bias units (which may be regarded as special neurons in form) may also be provided in any one or more layers in the long-short memory network model.
In an exemplary embodiment, the long-short memory network model may be used to predict the number of short periods that a user should take some action in the next M short periods, such as the number of days that coupons are taken in the next 7 days. The prediction result should be an integer between 0, M, and the user behavior prediction can be converted into a classification problem of m+1 results, so s=m+1 can be derived.
In an exemplary embodiment, the first fully-connected layer and the second fully-connected layer may be ReLu active layers and the output layer may be Softmax active layers. ReLu activation functions are used in the fully connected layer, which can be thinned while features are extracted. And the Softmax activation function is suitable for normalizing the output numerical value to obtain a classification result.
In an exemplary embodiment, the user behavior prediction method may refer to fig. 5, and includes the following steps:
Step S510, counting behavior data of a user to be predicted in the last N long periods, and converting the behavior data into N feature vectors with sequence;
Step S520, inputting the first feature vector into a long-short-time memory network model to obtain a first long-short-time memory vector and a first output vector G1;
Step S530, sequentially updating the initial memory state of the long-short-time memory network model according to the i-1 th long-short-time memory vector, and inputting the i-th feature vector into the long-short-time memory network model to obtain an i-th long-time memory vector and an i-th output vector G i, wherein i is [2, N-1];
step S540, updating the initial memory state of the long-short-time memory network model according to the N-1 long-short-time memory vector, and inputting the N characteristic vector into the long-short-time memory network model to obtain the N output vector;
Step S550, sequentially determining an ith real vector Hi according to the ith feature vector, and matching the ith real vector H i with the ith-1 output vector G i-1;
Step S560, if the following condition (1) is satisfied, it is determined that the current user behavior prediction is not valid:
wherein, K is a matching threshold, II (·) is an indication function, and values 1 and 0 are respectively taken when the·in the brackets is true and false;
In step S570, if the condition (1) is not satisfied, determining a behavior prediction result of the user to be predicted according to the nth output vector.
In the long-short-term memory network model, the output vector represents a prediction of the user behavior of the next long period, for example, the first output vector G1 corresponding to the first feature vector is a prediction result of the second long period. In the present exemplary embodiment, the first to nth long periods are the most recent N long periods, i.e., the history period, and the real behavior result may be obtained according to the second feature vector of the second long period, and may be represented by the real vector H2. Comparing H2 with G1, it can be known whether the prediction result of G1 is accurate. In the present exemplary embodiment, G N is an output vector that needs to be obtained finally to predict the behavior of the user in a long period in the future, and G1 to G N-1 may be regarded as intermediate data generated in the model processing process, which may reflect the accuracy of model prediction to a certain extent. Thus, H i and G i-1 are matched in sequence and the degree of statistical match is accumulated, usingCharacterization, the larger the value, the higher the model accuracy. A matching threshold K can be set as a judgment criterion ifThe method and the device indicate that the effect of the model in the previous N-1 round of prediction is not good, the result of the Nth round of prediction is not reliable, the user behavior prediction can be judged to be invalid, otherwise, the method and the device indicate that the effect of the model is good, and the final behavior prediction result can be determined through G N. The value of K may be empirically set, for example, k=0.7· (N-1), which indicates that the model requires at least 70% of the number of rounds of prediction to be accurate in the previous N-1 round of prediction, or may be optimized in the actual application process, for example, the effective user behavior prediction result may be compared with the subsequent actual user behavior, if there is more inconsistency, the value of K may be appropriately increased, the ineffective user behavior prediction result may be compared with the subsequent actual user behavior, if there is more consistency, the value of K may be appropriately decreased, etc., which is not particularly limited in the present disclosure.
In other embodiments, because the long-short-term memory network model has a time sequence accumulation effect, the larger the number of predicted rounds is, the more reliable the predicted result is, so that the matching condition of each round of predicted result can be weighted and counted, that is, when the following condition (2) is satisfied, the invalid prediction of the current user behavior can be determined:
wherein the weight of each round of prediction may be equal to the number of rounds. K in condition (2) is a different form of matching threshold than K in condition (1), which values are typically not equal, e.g., K in condition (1) is 0.7 (N-1), K in condition (2) may be 0.7.
The specific determination conditions to be employed are not particularly limited in the present disclosure.
The exemplary embodiment of the disclosure further provides a user behavior prediction apparatus, referring to fig. 6, the apparatus 600 may include a data acquisition module 610 configured to acquire behavior data of a user to be predicted in N long periods and convert the data into N feature vectors having a sequence, a model analysis module 620 configured to input a first feature vector into a long-short-time memory network model to obtain the first long-short-time memory vector, sequentially update an initial memory state of the long-short-time memory network model according to the i-1 th long-time memory vector and input the i feature vector into the long-short-time memory network model to obtain the i-long-time memory vector, and update the initial memory state of the long-short-time memory network model according to the N-1 th long-time memory vector and input the N-th feature vector into the long-short-time memory network model to obtain an output vector, and a result output module 630 configured to determine a behavior prediction result of the user to be predicted according to the output vector.
In an exemplary embodiment, the data obtaining module may be further configured to sequentially count O behavior data of the user to be predicted in the last N long periods according to the short periods, where each long period includes M short periods, and convert the behavior data into N m×o dimensional feature vectors with a sequence.
In an exemplary embodiment, the long-short memory network model at least comprises an input layer, a first full-connection layer, a long-short memory layer and a second full-connection layer, wherein the input layer comprises M & ltO & gt neurons and is used for inputting M & ltO & gt-dimensional feature vectors, the first full-connection layer comprises P neurons and is used for converting the feature vectors into P-dimensional first intermediate vectors, the long-short memory layer comprises Q neurons and is used for converting the first intermediate vectors into Q-dimensional long-short memory vectors according to initial memory states, any two neurons in the Q neurons are connected, the second full-connection layer comprises R neurons and is used for converting the long-short memory vectors into R-dimensional second intermediate vectors, and the output layer comprises S neurons and is used for converting the second intermediate vectors into S-dimensional output vectors.
In an exemplary embodiment, the long-short-time memory layer is a forward long-short-time memory layer, the long-short-time memory vector is a forward long-short-time memory vector, the initial memory state is a forward initial memory state, the long-short-time memory network model further comprises a backward long-time memory layer, the backward long-time memory layer comprises T neurons and is used for converting the forward long-time memory vector into a T-dimensional backward long-short-time memory vector according to the backward initial memory state, any two neurons in the T neurons are connected, the second full-connection layer is used for converting the backward long-time memory vector into a second intermediate vector, the model analysis module can be used for inputting the first characteristic vector into a long-short-time memory network model to obtain a first forward long-time memory vector, sequentially updating the forward initial memory state and the backward initial memory state of the long-time memory network model according to the i-1 forward long-time memory vector and the i-1 backward long-time memory vector respectively, and inputting the i characteristic vector into the long-short-time memory network model to obtain the i forward long-time memory vector and the i backward long-time memory model.
In an exemplary embodiment, s=m+1.
In an exemplary embodiment, the first fully-connected layer and the second fully-connected layer may be ReLu active layers and the output layer may be Softmax active layers.
In an exemplary embodiment, the model analysis module may be further configured to input a first feature vector into a long-short-time memory network model to obtain a first long-short-time memory vector and a first output vector, sequentially update an initial memory state of the long-short-time memory network model according to the i-1 th long-short-time memory vector, and input the i feature vector into the long-short-time memory network model to obtain an i-th long-short-time memory vector and an i-th output vector, where i is [2, N-1];
Wherein K is a matching threshold.
The specific details of the above modules are already described in the embodiments of the method section, and thus are not repeated.
The exemplary embodiments of the present disclosure also provide an electronic device capable of implementing the above method.
Those skilled in the art will appreciate that the various aspects of the present disclosure may be implemented as a system, method, or program product. Accordingly, aspects of the present disclosure may be embodied in the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, micro-code, etc.) or an embodiment combining hardware and software aspects that may be referred to herein collectively as a "circuit," module, "or" system.
An electronic device 700 according to such an exemplary embodiment of the present disclosure is described below with reference to fig. 7. The electronic device 700 shown in fig. 7 is merely an example and should not be construed to limit the functionality and scope of use of embodiments of the present disclosure in any way.
As shown in fig. 7, the electronic device 700 is embodied in the form of a general purpose computing device. The components of the electronic device 700 may include, but are not limited to, the at least one processing unit 710 described above, the at least one storage unit 720 described above, a bus 730 connecting the various system components (including the storage unit 720 and the processing unit 710), and a display unit 740.
Wherein the storage unit stores program code that is executable by the processing unit 710 such that the processing unit 710 performs steps according to various exemplary embodiments of the present disclosure described in the above-described "exemplary methods" section of the present specification. For example, the processing unit 710 may execute steps S110 to S150 shown in fig. 1, or may execute steps S510 to S570 shown in fig. 5.
The memory unit 720 may include readable media in the form of volatile memory units, such as Random Access Memory (RAM) 721 and/or cache memory 722, and may further include Read Only Memory (ROM) 723.
The storage unit 720 may also include a program/utility 724 having a set (at least one) of program modules 725, such program modules 725 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.
Bus 730 may be a bus representing one or more of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
The electronic device 700 may also communicate with one or more external devices 900 (e.g., keyboard, pointing device, bluetooth device, etc.), one or more devices that enable a user to interact with the electronic device 700, and/or any device (e.g., router, modem, etc.) that enables the electronic device 700 to communicate with one or more other computing devices. Such communication may occur through an input/output (I/O) interface 750. Also, electronic device 700 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through network adapter 760. As shown, network adapter 760 communicates with other modules of electronic device 700 over bus 730. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with electronic device 700, including, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
From the above description of embodiments, those skilled in the art will readily appreciate that the example embodiments described herein may be implemented in software, or may be implemented in software in combination with the necessary hardware. Thus, the technical solutions according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.) or on a network, including several instructions to cause a computing device (may be a personal computer, a server, a terminal device, or a network device, etc.) to perform the method according to the exemplary embodiments of the present disclosure.
Exemplary embodiments of the present disclosure also provide a computer readable storage medium having stored thereon a program product capable of implementing the method described above in the present specification. In some possible implementations, various aspects of the disclosure may also be implemented in the form of a program product comprising program code for causing a terminal device to carry out the steps according to the various exemplary embodiments of the disclosure as described in the "exemplary methods" section of this specification, when the program product is run on the terminal device.
Referring to fig. 8, a program product 800 for implementing the above-described method according to an exemplary embodiment of the present disclosure is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of a readable storage medium include an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The computer readable signal medium may include a data signal propagated in baseband or as part of a carrier wave with readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Program code for carrying out operations of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., connected via the Internet using an Internet service provider).
Furthermore, the above-described figures are only schematic illustrations of processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It will be readily appreciated that the processes shown in the above figures do not indicate or limit the temporal order of these processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, for example, among a plurality of modules.
It should be noted that although in the above detailed description several modules or units of a device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functionality of two or more modules or units described above may be embodied in one module or unit in accordance with exemplary embodiments of the present disclosure. Conversely, the features and functions of one module or unit described above may be further divided into a plurality of modules or units to be embodied.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any adaptations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It is to be understood that the present disclosure is not limited to the precise arrangements and instrumentalities shown in the drawings, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims (10)

1.一种用户行为预测方法，其特征在于，包括：1. A user behavior prediction method, comprising: 获取待预测用户在最近N个长周期内的行为数据，并转换为具有先后顺序的N个特征向量；Obtain the behavior data of the user to be predicted in the last N long periods and convert it into N feature vectors with a chronological order; 将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量；Inputting the first feature vector into a long short-term memory network model to obtain a first long short-term memory vector; 依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量，其中i∈[2,N-1]；Sequentially, the initial memory state of the long short-term memory network model is updated according to the i-1th long short-term memory vector, and the i-th feature vector is input into the long short-term memory network model to obtain the i-th long short-term memory vector, where i∈[2,N-1]; 根据第N-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第N特征向量输入所述长短时记忆网络模型，得到输出向量；Updating the initial memory state of the long short-term memory network model according to the N-1th long short-term memory vector, and inputting the Nth feature vector into the long short-term memory network model to obtain an output vector; 根据所述输出向量确定所述待预测用户的行为预测结果；Determine the behavior prediction result of the to-be-predicted user according to the output vector; 其中，所述将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量包括：将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量与第一输出向量；The step of inputting the first feature vector into a long short-term memory network model to obtain the first long short-term memory vector comprises: inputting the first feature vector into a long short-term memory network model to obtain the first long short-term memory vector and the first output vector; 所述依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量包括：依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量与第i输出向量；The step of sequentially updating the initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector comprises: sequentially updating the initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector and the i-th output vector; 所述方法还包括：依次根据所述第i特征向量确定第i真实向量Hi，并将所述第i真实向量Hi与第i-1输出向量Gi-1进行匹配，通过累积统计匹配程度得到统计值；若统计值小于匹配阈值，则判定本次用户行为预测无效，若统计值不小于匹配阈值，则执行根据所述输出向量确定所述待预测用户的行为预测结果的步骤，其中所述输出向量为第N输出向量。The method also includes: determining the i-th true vector Hi according to the i-th feature vector in turn, and matching the i-th true vector Hi with the i-1-th output vector Gi -1 , and obtaining a statistical value by accumulating statistical matching degrees; if the statistical value is less than a matching threshold, determining that the current user behavior prediction is invalid; if the statistical value is not less than the matching threshold, executing the step of determining the behavior prediction result of the user to be predicted according to the output vector, wherein the output vector is the N-th output vector. 2.根据权利要求1所述的方法，其特征在于，所述获取待预测用户在最近N个长周期内的行为数据，并转换为具有先后顺序的N个特征向量包括：2. The method according to claim 1 is characterized in that the step of obtaining the behavior data of the user to be predicted in the most recent N long periods and converting the data into N feature vectors in a sequential order comprises: 按照短周期依次统计待预测用户在最近N个长周期内的O种行为数据，每个长周期包括M个短周期；According to the short cycle, the O types of behavior data of the user to be predicted in the latest N long cycles are counted in sequence, and each long cycle includes M short cycles; 将所述行为数据转换为具有先后顺序的N个M*O维特征向量。The behavior data is converted into N M*O dimensional feature vectors with a sequence. 3.根据权利要求2所述的方法，其特征在于，所述长短时记忆网络模型至少包括：3. The method according to claim 2, characterized in that the long short-term memory network model at least includes: 输入层，包括M*O个神经元，用于输入所述M*O维特征向量；An input layer, comprising M*O neurons, for inputting the M*O dimensional feature vector; 第一全连接层，包括P个神经元，用于将所述特征向量转换为P维的第一中间向量；A first fully connected layer, including P neurons, is used to convert the feature vector into a first intermediate vector of P dimension; 长短时记忆层，包括Q个神经元，用于根据所述初始记忆状态将所述第一中间向量转换为Q维的长短时记忆向量，所述Q个神经元中任意两个神经元相连接；a long short-term memory layer, comprising Q neurons, configured to convert the first intermediate vector into a Q-dimensional long short-term memory vector according to the initial memory state, wherein any two neurons among the Q neurons are connected; 第二全连接层，包括R个神经元，用于将所述长短时记忆向量转换为R维的第二中间向量；A second fully connected layer, including R neurons, is used to convert the long short-term memory vector into a second intermediate vector of R dimension; 输出层，包括S个神经元，用于将所述第二中间向量转换为S维的输出向量。The output layer includes S neurons and is used to convert the second intermediate vector into an S-dimensional output vector. 4.根据权利要求3所述的方法，其特征在于，所述长短时记忆层为前向长短时记忆层，所述长短时记忆向量为前向长短时记忆向量，所述初始记忆状态为前向初始记忆状态；所述长短时记忆网络模型还包括：4. The method according to claim 3, characterized in that the long short-term memory layer is a forward long short-term memory layer, the long short-term memory vector is a forward long short-term memory vector, and the initial memory state is a forward initial memory state; the long short-term memory network model further comprises: 后向长短时记忆层，包括T个神经元，用于根据后向初始记忆状态将所述前向长短时记忆向量转换为T维的后向长短时记忆向量，所述T个神经元中任意两个神经元相连接；A backward long short-term memory layer, comprising T neurons, for converting the forward long short-term memory vector into a T-dimensional backward long short-term memory vector according to a backward initial memory state, wherein any two neurons among the T neurons are connected; 则所述第二全连接层用于将所述后向长短时记忆向量转换为所述第二中间向量；Then the second fully connected layer is used to convert the backward long short-term memory vector into the second intermediate vector; 所述将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量包括：The step of inputting the first feature vector into a long short-term memory network model to obtain the first long short-term memory vector comprises: 将第一特征向量输入一长短时记忆网络模型，得到第一前向长短时记忆向量；Inputting the first feature vector into a long short-term memory network model to obtain a first forward long short-term memory vector; 所述依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量包括：The step of sequentially updating the initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector comprises: 依次地，分别根据第i-1前向长短时记忆向量与第i-1后向长短时记忆向量更新所述长短时记忆网络模型的前向初始记忆状态与后向初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i前向长短时记忆向量与第i后向长短时记忆向量。Sequentially, the forward initial memory state and the backward initial memory state of the long short-term memory network model are updated according to the i-1th forward long short-term memory vector and the i-1th backward long short-term memory vector, and the i-th feature vector is input into the long short-term memory network model to obtain the i-th forward long short-term memory vector and the i-th backward long short-term memory vector. 5.根据权利要求3所述的方法，其特征在于，S＝M+1。The method according to claim 3 , characterized in that S=M+1. 6.根据权利要求3所述的方法，其特征在于，所述第一全连接层与第二全连接层为线性修正单元ReLu激活层；所述输出层为归一化指数函数Softmax激活层。6. The method according to claim 3 is characterized in that the first fully connected layer and the second fully connected layer are linear rectified unit ReLu activation layers; and the output layer is a normalized exponential function Softmax activation layer. 7.根据权利要求1所述的方法，其特征在于，所述统计值小于匹配阈值，包括以下至少一种情况：7. The method according to claim 1, wherein the statistical value is less than the matching threshold, including at least one of the following situations: 其中，K为匹配阈值。Among them, K is the matching threshold. 8.一种用户行为预测装置，其特征在于，包括：8. A user behavior prediction device, comprising: 数据获取模块，用于获取待预测用户在N个长周期内的行为数据，并转换为具有先后顺序的N个特征向量；The data acquisition module is used to obtain the behavior data of the user to be predicted in N long periods and convert it into N feature vectors with a chronological order; 模型分析模块，用于将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量，依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量，以及根据第N-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第N特征向量输入所述长短时记忆网络模型，得到输出向量，其中i∈[2,N-1]；A model analysis module, used for inputting a first feature vector into a long short-term memory network model to obtain a first long short-term memory vector, sequentially, updating an initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector, and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector, and updating the initial memory state of the long short-term memory network model according to the N-1th long short-term memory vector, and inputting the N-th feature vector into the long short-term memory network model to obtain an output vector, wherein i∈[2,N-1]; 结果输出模块，用于根据所述输出向量确定所述待预测用户的行为预测结果；A result output module, used to determine the behavior prediction result of the user to be predicted according to the output vector; 其中，所述将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量包括：将第一特征向量输入一长短时记忆网络模型，得到第一长短时记忆向量与第一输出向量；The step of inputting the first feature vector into a long short-term memory network model to obtain the first long short-term memory vector comprises: inputting the first feature vector into a long short-term memory network model to obtain the first long short-term memory vector and the first output vector; 所述依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量包括：依次地，根据第i-1长短时记忆向量更新所述长短时记忆网络模型的初始记忆状态，并将第i特征向量输入所述长短时记忆网络模型，得到第i长短时记忆向量与第i输出向量；The step of sequentially updating the initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector comprises: sequentially updating the initial memory state of the long short-term memory network model according to the i-1th long short-term memory vector and inputting the i-th feature vector into the long short-term memory network model to obtain the i-th long short-term memory vector and the i-th output vector; 所述结果输出模块还用于：依次根据所述第i特征向量确定第i真实向量Hi，并将所述第i真实向量Hi与第i-1输出向量Gi-1进行匹配，通过累积统计匹配程度得到统计值；若统计值小于匹配阈值，则判定本次用户行为预测无效，若统计值不小于匹配阈值，则执行根据所述输出向量确定所述待预测用户的行为预测结果的步骤，其中所述输出向量为第N输出向量。The result output module is also used to: determine the i-th real vector Hi according to the i-th feature vector in turn, and match the i-th real vector Hi with the i-1-th output vector Gi -1 , and obtain a statistical value by accumulating statistical matching degrees; if the statistical value is less than a matching threshold, it is determined that the user behavior prediction is invalid; if the statistical value is not less than the matching threshold, the step of determining the behavior prediction result of the user to be predicted according to the output vector is executed, wherein the output vector is the N-th output vector. 9.一种电子设备，其特征在于，包括：9. An electronic device, comprising: 处理器；以及Processor; and 存储器，用于存储所述处理器的可执行指令；A memory, configured to store executable instructions of the processor; 其中，所述处理器配置为经由执行所述可执行指令来执行权利要求1-7任一项所述的方法。The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions. 10.一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现权利要求1-7任一项所述的方法。10. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.
CN201811428883.0A 2018-11-27 2018-11-27 User behavior prediction method and device, electronic equipment and storage medium Active CN111221896B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811428883.0A CN111221896B (en) 2018-11-27 2018-11-27 User behavior prediction method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811428883.0A CN111221896B (en) 2018-11-27 2018-11-27 User behavior prediction method and device, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN111221896A CN111221896A (en) 2020-06-02
CN111221896B true CN111221896B (en) 2025-01-17

Family

ID=70830480

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811428883.0A Active CN111221896B (en) 2018-11-27 2018-11-27 User behavior prediction method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN111221896B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112275438B (en) * 2020-10-13 2022-03-01 成都智叟智能科技有限公司 Control method and system for separation and crushing of wet and dry waste based on data analysis
CN112398663A (en) * 2020-11-06 2021-02-23 浪潮云信息技术股份公司 Elastic IP charging method and system based on deep neural network

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106682220A (en) * 2017-01-04 2017-05-17 华南理工大学 Online traditional Chinese medicine text named entity identifying method based on deep learning
CN107730087A (en) * 2017-09-20 2018-02-23 平安科技（深圳）有限公司 Forecast model training method, data monitoring method, device, equipment and medium
CN108305094A (en) * 2017-12-18 2018-07-20 北京三快在线科技有限公司 A kind of user's behavior prediction method and device, electronic equipment

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10002322B1 (en) * 2017-04-06 2018-06-19 The Boston Consulting Group, Inc. Systems and methods for predicting transactions
CN108829737B (en) * 2018-05-21 2021-11-05 浙江大学 Text cross-combination classification method based on bidirectional long short-term memory network

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106682220A (en) * 2017-01-04 2017-05-17 华南理工大学 Online traditional Chinese medicine text named entity identifying method based on deep learning
CN107730087A (en) * 2017-09-20 2018-02-23 平安科技（深圳）有限公司 Forecast model training method, data monitoring method, device, equipment and medium
CN108305094A (en) * 2017-12-18 2018-07-20 北京三快在线科技有限公司 A kind of user's behavior prediction method and device, electronic equipment

Also Published As

Publication number Publication date
CN111221896A (en) 2020-06-02

Similar Documents

Publication Publication Date Title
CN112118143B (en) Traffic prediction model training method, traffic prediction method, device, equipment and medium
WO2018171533A1 (en) Review machine learning system
CN111258593B (en) Application program prediction model building method and device, storage medium and terminal
CN112202726B (en) System anomaly detection method based on context sensing
US11811708B2 (en) Systems and methods for generating dynamic conversational responses using cluster-level collaborative filtering matrices
CN111611478B (en) Information recommendation method and device and electronic equipment
US20230305904A1 (en) Systems and methods for correlating probability models with non-homogenous time dependencies to generate time-specific data processing predictions
CN112508265A (en) Time and activity multi-task prediction method and system for business process management
CN114549001B (en) Training risky transaction identification model, method and device for identifying risky transactions
WO2021012263A1 (en) Systems and methods for end-to-end deep reinforcement learning based coreference resolution
CN114118570A (en) Service data prediction method and device, electronic equipment and storage medium
CN119047651B (en) Community energy consumption prediction method, device and equipment based on Bayesian calibration
CN111221896B (en) User behavior prediction method and device, electronic equipment and storage medium
CN119961493B (en) An intelligent data mining platform and method based on big data analysis
CN114120180A (en) Method, device, equipment and medium for generating time sequence nomination
CN119047116A (en) Method and system for predicting service life of cross-working-condition rolling bearing
CN109272165B (en) Registration probability estimation method and device, storage medium and electronic equipment
US11782812B2 (en) Causal attention-based multi-stream RNN for computer system metric prediction and influential events identification based on metric and event logs
CN119806937A (en) Anomaly detection method and related equipment for application service reliability
CN118114275B (en) A flexible permission decision method and device for multiple resources
CN113011674A (en) Photovoltaic power generation prediction method and device, electronic equipment and storage medium
CN113886821A (en) Malicious process identification method and device based on twin network, electronic equipment and storage medium
US11221934B2 (en) Identifying anomalies in data during data outage
CN118945270A (en) Call strategy method, device, equipment and medium based on predicting user behavior
US20240328410A1 (en) Artificial Intelligence for Sump Pump Monitoring and Service Provider Notification

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载