Gaussian Naive Bayes Classifier

• Gaussian Naive Bayes Classifier: Iris data set

  Nice write-up of applying the sklearn Naive_Bayes_Classifier to the Iris dataset.

Linear Methods

• PyData talk: Winning with Simple, even Linear, Models

  Great talk about how to get a lot of mileage with tricks using logistic and linear regression.

• Solving least squares by Matrix inversion

  Most ML libraries find the minima of e.g. a linear or logistic regression by means of gradient descent. However, you can find the minimum exactly by solving what is called in Linear algebra a Normal Equation

Regularization

• Ridge vs LASSO

  A nice plot giving intuition why Ridge (L²) regularization tries to make parameters small whereas LASSO (L¹) regularization tends to force some parameters towards zero, while others can stay large.

Encoding Data for ML Models

• Smarter Ways to Encode Categorical Data for Machine Learning

  Better encoding of categorical data can mean better model performance. In this article I’ll introduce you to a wide range of encoding options from the Category Encoders package for use with scikit-learn machine learning in Python.

Dimensionality reduction

• Singular Value Decomposition

  Good clear video explaining SVD from a Stanford AI course. (13 minutes)

• StatQuest: Principal Component Analysis (PCA), Step-by-Step

  YouTube video showing how to use SVD to perform PCA. (21 minutes)

---