clemtarge / clemtarge

FastML

FastML allows you to create, train, test, and save your Machine Learning model with two lines of code.
You can also optimize it with Bayesian Optimization by specifying model hyperparameters.

Reinforcement learning

Some codes to use Q-Learning or Actor-Critic agent in Pytorch. Actor-Critic is also available in Tensorflow 2.
The notebooks can be used as examples or as starting points.

Preprocessing (Python notebook)

In this notebook, you will find some functions to visualize, transform and replace NaN values in the data.

Preprocessing time series (Python notebook)

Some useful functions to deal with time series. Download climate/ for toydata.

Statistical tests (R notebook)

In this notebook, you will find some non parametric statistical tests to test:

• Adequation to a distribution
  • Kolmogorov test
  • Chi2 test
• Adequation to a normal distribution
  • Shapiro-Wilk test
  • Lilliefors test
  • Other methods
• Homogeneity between two samples
  • Wilcoxon-Mann-Whitney test
  • Kolmogorov-Smirnov test
  • Chi2 test
• Independance between two samples
  • Chi2 test