Weekly summary 2016-11-19

HMM/MDP project

• Lit review
  • [TLF] Monte Carlo HiddenMarkov Models: Learning Non-Parametric Models of Partially Observable Stochastic Processes (Mon.)
• Theory review
  • Ch. 3, 5 of FA
  • Ch. 6 Bertsekas
• Experiments
  • Program basic RL
• Theory: things to try
  • Nonparametric/kernel methods (cf. BCO)

Alexa/NLP experiments

• Get LSTM up and running.
• Get Yingyu’s word2vec, DL code up and running.
  • Try modifying with other modalities (sentiments, etc.)
• Lisa Lee’s thesis

Other experiments

• Neural net learns dictionaries: run experiments.
  • NNDL experiment.
  • Where is the cusp?

Learning

• Sum of squares: tensor decomposition (Mon.)
• EM algorithm
• Tensor decomposition and latent variable model
• MDP theory (see above)
• Summarize papers