RFX-Fuse: Breiman and Cutler's Unified ML Engine + Native Explainable Similarity
This addresses the problem of fragmented ML workflows for practitioners by providing a more integrated and efficient solution, though it is incremental as it revives and extends an existing vision.
The paper tackles the fragmentation of modern machine learning pipelines by introducing RFX-Fuse, a unified engine that implements Breiman and Cutler's original Random Forest vision, including classification, regression, unsupervised learning, similarity, outlier detection, imputation, and visualization in a single model object, achieving a 1 to 2 model alternative compared to 5+ separate tools.
Breiman and Cutler's original Random Forest was designed as a unified ML engine -- not merely an ensemble predictor. Their implementation included classification, regression, unsupervised learning, proximity-based similarity, outlier detection, missing value imputation, and visualization -- capabilities that modern libraries like scikit-learn never implemented. RFX-Fuse (Random Forests X [X=compression] -- Forest Unified Learning and Similarity Engine) delivers Breiman and Cutler's complete vision with native GPU/CPU support. Modern ML pipelines require 5+ separate tools -- XGBoost for prediction, FAISS for similarity, SHAP for explanations, Isolation Forest for outliers, custom code for importance. RFX-Fuse provides a 1 to 2 model object alternative -- a single set of trees grown once. Novel Contributions: (1) Proximity Importance -- native explainable similarity: proximity measures that samples are similar; proximity importance explains why. (2) Dataset-specific imputation validation for general tabular data -- ranking imputation methods by how real the imputed data looks, without ground truth labels.