BMRM Open source and scalable convex solver for many machine learning problems
Download

BMRM Ranking & Summary

BMRM Tags

solver machine learning convex solver machine learning solver solve minimization problem convex

BMRM Description

Open source and scalable convex solver for many machine learning problems BMRM is a modular, open source, and scalable convex solver for many machine learning problems cast in the form of regularized risk minimization problem.BMRM is "modular" because the (problem-specific) loss function module is decoupled from the (regularization-specific) optimization module (e.g. quadratic programming or linear programming solvers), thus shorten the time to implement/prototype solutions to new problems. Besides, the decoupling leads to easier parallelization of the loss function computation. Here are some key features of "BMRM": · Binary classification: Hinge, Squared hinge, Huber-hinge, Logistic regression, Exponential, ROC Score, Fbeta Score · Univariate regression: $epsilon$-insensitive, Huber robust, Least Mean Squares, Least Absolute Deviation · Novelty detection (1-class SVM) · Quantile regression · Poisson regression · Ranking: NDCG (normalized discounted cummulative gain) · Graph Matching · Sequence Segmentation and Classification

BMRM Related Software