Random Forest

Random Forest is a fast, fully offline app that grows a real random forest of classification trees from your own labelled 2-D points - each tree votes, and the majority wins. Watch the ensemble's blended decision regions form as trees agree or disagree, check the out-of-bag accuracy estimate, and see exactly which feature drives the predictions - all rendered live on your device. WHAT'S INSIDE Forest The signature view - blended decision regions shaded by how strongly the trees agree, with misclassified rows ringed, a plain-language verdict, and an overfitting check. Data Steppers for tree count and per-tree depth, plus an editable table of (x, y, class) rows. Out-of-bag and training accuracy, feature importances, and class sizes update live. Importance See the Gini-decrease importance of each feature as simple bars, so you know exactly which variable is doing the work. Concept A plain-English walkthrough of the random forest idea - wisdom of the crowd, bagging, random feature selection, and out-of-bag accuracy - with the vote and importance formulas in offline KaTeX. Simulate Knobs for classes, rows per class, tree count and overlap. Generate noisy clouds and watch the out-of-bag accuracy steady as the forest grows. Examples Ready-made datasets - XOR, three clusters, noisy overlap, and an x-dominant set - that load with one tap. WHY RANDOM FOREST - 100% offline - your data never leaves your device - No account, no sign-up, no clutter - Clean, distinctive interface built for students, researchers, and anyone learning ensemble methods - Bagging, random feature selection, out-of-bag accuracy, and Gini importance - the real algorithm, not a toy - Input is never lost - trees, depth, and every row are saved automatically Whether you're a student learning why random forests beat a single decision tree, a researcher sanity-checking a small classification problem, or just curious how ensemble voting works, Random Forest shows you the whole mechanism in one glance.