/ml

Implementations of algorithms described in Pattern Recognition and Machine Learning

Primary LanguageJupyter Notebook

TODO

  • build helper functions
    • grid search with specified values
    • exponential grid search
    • make class interfaces work well with grid search in a general way (common interface for hyperparameters)
  • functions to evaluate model performance
  • functions to output a grid of graphs from grid search for visualization
  • visualize the performance of a model as data size increases
  • visualize prior / likelihood / model after each batch update like in PRML