ML: evaluation and optimization

Twelve experiments in honest preprocessing, metrics, validation, optimization and unsupervised learning.

Progress and your code are kept in this browser only.

1. 1Train-only Standardization
2. 2Confusion Matrix
3. 3Precision, Recall & F1
4. 4Threshold Sweep
5. 5Cross-validation Folds
6. 6Ridge Regression
7. 7Numerical Gradient Check
8. 8Adam Optimizer
9. 9Mini-batch SGD
10. 10Multinomial Naive Bayes
11. 11PCA by Power Iteration
12. 12Silhouette Clustering Score