/Statistical-Optimization

Statistical optimization for AI and machine learning

Primary LanguagePython

Some material from my book Statistical Optimization for AI and Machine Learning, available here. In particular:

  • My gradient descent technique implemented in gradient.py, available in this folder.
  • The interpol.py, interpol_fourier.py and interpol_ortho.py programs in this folder are described in my article New Interpolation Methods for Data Synthetization and Prediction, available here.
  • For feature clustering, see featureClustering.py and featureClusteringScipy.py (the latter with hierarchical clustering) in this folder.
  • Fast grid search for faster hyperparameter tuning: see ZetaGeom.py in this folder. The article describing and documenting the method is available here.
  • Stochastic thinning: new technique to boost learning algorithms. See thinning_neuralNets.py, and thinning_regression.py in this folder. The article describing and documenting the method is available here.
  • Extrapolated quantiles (quantile convolution) to debias GenAI methods. See equantile.py. The article describing and documenting the method is available here.
  • Material about Generative Adversarial Networks (GAN), NoGAN and NoGAN2, is in the main folder. The corresponding Python libraries are described in the book.