#machine-learning #svm #libsvm

nightly ffsvm

A libSVM compatible support vector machine, but up to 10x faster, for games or VR

5 unstable releases

0.6.0 Aug 15, 2018
0.5.0 Aug 9, 2018
0.4.3 Aug 5, 2018
0.4.1 Aug 5, 2018
0.4.0 Aug 5, 2018

#16 in Machine learning

Download history 72/week @ 2018-08-11 24/week @ 2018-08-18 12/week @ 2018-08-25 11/week @ 2018-09-08 15/week @ 2018-09-15 4/week @ 2018-09-22 18/week @ 2018-09-29 6/week @ 2018-10-06 12/week @ 2018-10-13 98/week @ 2018-10-20 5/week @ 2018-10-27

31 downloads per month

MIT license

82KB
1.5K SLoC

Latest Version Travis-CI Status docs MIT

In One Sentence

You trained a SVM using libSVM, now you want the highest possible performance during (real-time) classification, like games or VR.

Highlights

  • loads almost all libSVM types (C-SVC, ν-SVC, ε-SVR, ν-SVR) and kernels (linear, poly, RBF and sigmoid)
  • produces practically same classification results as libSVM
  • optimized for SIMD and can be mixed seamlessly with Rayon
  • written in 100% Rust, but can be loaded from any language (via FFI)
  • allocation-free during classification for dense SVMs
  • 2.5x - 14x faster than libSVM for dense SVMs
  • extremely low classification times for small models (e.g., 128 SV, 16 dense attributes, linear ~ 500ns)
  • successfully used in Unity and VR projects (Windows & Android)
  • free of unsafe code ;)

Usage

Train with libSVM (e.g., using the tool svm-train), then classify with ffsvm-rust.

From Rust:

// Replace `SAMPLE_MODEL` with a `&str` to your model.
let svm = DenseSVM::try_from(SAMPLE_MODEL)?;

let mut problem = Problem::from(&svm);
let features = problem.features();

features[0] = 0.55838;
features[1] = -0.157895;
features[2] = 0.581292;
features[3] = -0.221184;

svm.predict_value(&mut problem)?;

assert_eq!(problem.solution(), Solution::Label(42));

From C / FFI:

Please see FFSVM-FFI

Status

  • Aug 5, 2018: Still in alpha, but finally on crates.io.
  • May 27, 2018: We're in alpha. Successfully used internally on Windows, Mac, Android and Linux on various machines and devices. Once SIMD stabilizes and we can cross-compile to WASM we'll move to beta.
  • December 16, 2017: We're in pre-alpha. It will probably not even work on your machine.

Performance

performance

Classification time vs. libSVM for dense models.

performance

Performance milestones during development.

All performance numbers reported for the DenseSVM. We also have support for SparseSVMs, which are slower for "mostly dense" models, and faster for "mostly sparse" models (and generally on the performance level of libSVM).

See here for details.

FAQ

See here for details.

Dependencies

~3MB
~61K SLoC