Repository files navigation

Port of libsvm v3.22 usingemscripten and it targets browser or nodejs using asm and WebAssembly.

What is libsvm?libsvm is ac++ library developped by Chih-Chung Chang and Chih-Jen Lin that allows to do support vector machine (aka SVM) classification and regression.

Resources about libsvm:

• libsvm website
• libsvm github repository
• Wikipedia article
• Chih-Chung Chang and Chih-Jen Lin, LIBSVM : a library for support vector machines. ACM Transactions on Intelligent Systems and Technology, 2:27:1--27:27, 2011.

npm install libsvm-ts

yarn add libsvm-ts

The main entry point loads the WebAssembly build and is asynchronous.

import{SVM}from'libsvm-ts';constsvm=newSVM();// ...

The npm package contains a bundle for the browser that works with AMD and browser globals. There is one bundle for the asm build and another for the web assembly build. They are located in thedist/browser directory of the package. You can load them into your web page with ascript tag. For the web assembly module, make sure that the libsvm.wasm file is served from the same relative path as the js file.

This example illustrates how to use the library to train and use an SVM classifier.

import{SVM}from'libsvm-ts';constsvm=newSVM({type:'C_SVC',kernel:'RBF',gamma:1,cost:1});// This is the xor problem//  1  0//  0  1constsamples=[[0,0],[1,1],[1,0],[0,1]];constlabels=[0,0,1,1];svm.loadWASM().then((loadedSVM)=>{// Also you can loadASM() instead of loadWASM()loadedSVM.train({ samples, labels});constpredictedLabel=loadedSVM.predictOne({sample:[0.7,0.8]});console.log(predictedLabel)// 0});

We welcome new contributors of all level of experience. The development guide will be added to assist new contributors to easily join the project.

• You want to participate in a Machine Learning project, which will boost your Machine Learning skills and knowledge
• Looking to be part of a growing community
• You want to learn Machine Learning
• You like Typescript ❤️ Machine Learning

CheckoutCONTRIBUTIONS.md for more information

You can compare the performance of the library in various environments. Runnpm run benchmark to run the benchmarks with native c/c++ code and with the compiled code with your local version of node.js. For browser performance, go to theweb benchmark page.

Speed is mainly affected by the javascript engine that compiles it. Since WebAssembly has been stabilized and is an optimization phase, more recent engines are almost always faster.

Speed is also affected by the version of emscripten that generated the build or the options used in the build. I will try to keep up with any improvement that might significantly impact the performance.

I report the results here for the cross-validation benchmark on the iris dataset to get a feeling for how performance compares on different platforms. There are other benchmarks that can be run from the terminal in node.js or in the browser. The performance results are given relative to how they run natively (with compiled c++ code). The benchmarks only consider runtime performance, not load and parse performance.

Fromasmjs.org

asm is an optimizable subset of javascript.

Fromwebassembly.org

WebAssembly or wasm is a new portable, size- and load-time-efficient format suitable for compilation to the web

Both. You should try to use WebAssembly first and fall back to asm in order to support all browsers.

WebAssembly is currently supported in the latest stable versions of Chrome, Firefox and on preview versions of Safari and Edge.

Kind: global class

• SVM
  • new SVM(options)
  • instance
    • .train(samples, labels)
    • .crossValidation(samples, labels, kFold) ⇒Array.<number>
    • .free()
    • .predictOne(sample) ⇒number
    • .predict(samples) ⇒Array.<number>
    • .predictProbability(samples) ⇒Array.<object>
    • .predictOneProbability(sample) ⇒object
    • .predictOneInterval(sample, confidence) ⇒object
    • .predictInterval(samples, confidence) ⇒Array.<object>
    • .getLabels() ⇒Array.<number>
    • .getSVIndices() ⇒Array.<number>
    • .serializeModel() ⇒string
  • static
    • .SVM_TYPES :Object
    • .KERNEL_TYPES :Object
    • .load(serializedModel) ⇒SVM

Trains the SVM model.

Kind: instance method ofSVM
Throws:

• if SVM instance was instantiated from SVM.load.

Performs k-fold cross-validation (KF-CV). KF-CV separates the data-set into kFold random equally sized partitions,and uses each as a validation set, with all other partitions used in the training set. Observations left overfrom if kFold does not divide the number of observations are left out of the cross-validation process. IfkFold is one, this is equivalent to a leave-on-out cross-validation

Kind: instance method ofSVM
Returns:Array.<number> - The array of predicted labels produced by the cross validation. Has a size equal tothe number of samples provided as input.
Throws:

• if SVM instance was instantiated from SVM.loadWASM() or SVM.loadASK().

Free the memory allocated for the model. Since this memory is stored in the memory model of emscripten, it isallocated within an ArrayBuffer and WILL NOT BE GARBARGE COLLECTED, you have to explicitly free it. Sonot calling this will result in memory leaks. As of today in the browser, there is no way to hook thegarbage collection of the SVM object to free it automatically.Free the memory that was created by the compiled libsvm library to.store the model. This model is reused every time the predict method is called.

Kind: instance method ofSVM

Predict the label of one sample.

Kind: instance method ofSVM
Returns:number - - The predicted label.

Predict the label of many samples.

Kind: instance method ofSVM
Returns:Array.<number> - - The predicted labels.

Predict the label with probability estimate of many samples.

Kind: instance method ofSVM
Returns:Array.<object> - - An array of objects containing the prediction label and the probability estimates for each label

Predict the label with probability estimate.

Kind: instance method ofSVM
Returns:object - - An object containing the prediction label and the probability estimates for each label

Predict a regression value with a confidence interval

Kind: instance method ofSVM
Returns:object - - An object containing the prediction value and the lower and upper bounds of the confidence interval

Predict regression values with confidence intervals

Kind: instance method ofSVM
Returns:Array.<object> - - An array of objects each containing the prediction label and the probability estimates for each label

Get the array of labels from the model. Useful when creating an SVM instance with SVM.load

Kind: instance method ofSVM
Returns:Array.<number> - - The list of labels.

Get the indices of the support vectors from the training set passed to the train method.

Kind: instance method ofSVM
Returns:Array.<number> - - The list of indices from the training samples.

Uses libsvm's serialization method of the model.

Kind: instance method ofSVM
Returns:string - The serialization string.

SVM classification and regression types

Kind: static property ofSVM
Properties

SVM kernel types

Kind: static property ofSVM
Properties

Create a SVM instance using ASM.

Kind: a member method ofSVM
Returns:SVM - - SVM instance that contains the model.

Create a SVM instance using WASM.

Kind: a member method ofSVM
Returns:SVM - - SVM instance that contains the model.

BSD-3-Clause