linfa

Why we included this project

Rust teams who want to keep their stack in a single language have a solid option in Linfa, a machine learning toolkit that plays the role scikit-learn does for Python. It is built as a set of small crates, each holding one algorithm or a tight family of them, covering the workhorse methods like linear and logistic regression, SVM, decision trees, k-means, DBSCAN, and naive Bayes, along with preprocessing helpers for normalization, TF-IDF vectorization, and PCA. Because the algorithms are split into separate sub-crates, a project can pull in just what it needs without adopting a heavy dependency. The same modularity keeps the code readable, which helps when your team wants to understand and audit what a model actually computes. If introducing another language or runtime was the thing holding you back from doing ML, Linfa keeps that work inside Rust.