Blog: http://fseoane.net/blog/

Things I'm working on:

• Manifold learning (and k-nearest neighbors)

Future work (hopefully before June 2011):

• Implement active sets (KKT conditions) in coordinate descent implementations for Lasso and ElasticNet. The ultimate goal is to make Lasso and ElasticNet comparable in all cases with the equivalents in glmnet.
• Work some more in the API of Gaussian Mixture Models
• Implement Affinity Propagation for sparse data.
• See how Spectral clustering could benefit from function implemented in pyamg