Logistic Regression

Logistic regression is a widely used discriminative classification model \(\mathbb{P}(y \mid \boldsymbol{x} ; \boldsymbol{\theta})\), where \(\boldsymbol{x} \in \mathbb{R}^D\) is a fixed-dimensional input vector, \(y \in\{1, \ldots, K\}\) is the class label, and \(\boldsymbol{\theta}\) are the parameters. If \(K=2\), this is known as binary logistic regression, and if \(K>2\), it is known as multinomial logistic regression, or alternatively, multiclass logistic regression [Murphy, 2019].

Further Readings

• https://d2l.ai/

• Murphy, Kevin P. “Chapter .” In Probabilistic Machine Learning: An Introduction. MIT Press, 2022.

• James, Gareth, Daniela Witten, Trevor Hastie, and Robert Tibshirani. “Chapter .” In An Introduction to Statistical Learning: With Applications in R. Boston: Springer, 2022.

• Jung, Alexander. “Chapter ” In Machine Learning: The Basics. Singapore: Springer Nature Singapore, 2023.

• Bishop, Christopher M. “Chapter .” In Pattern Recognition and Machine Learning. New York: Springer-Verlag, 2016.

• Hal Daumé III. “Chapter .” In A Course in Machine Learning, January 2017.

• Machine Learning from Scratch

• GOOD: https://github.com/NathanielDake/intuitiveml

• https://github.com/goodboychan/goodboychan.github.io/tree/main/_notebooks