Packages

esvm

1.1.0
1.1.1 1.1.0 1.0.0

An erlang library for SVM classification and regression based on libsvm.

Current section

Files

Jump to
Readme
3 Versions
0 Dependencies
Dependants
Files
Activity
esvm src esvm.erl
Raw

src/esvm.erl 

-module(esvm).



-include("esvm.hrl").



-export([

    model_create/3,

    model_load/1,

    model_save/2,

    model_predict/2

]).



-spec model_create([{label(), features()}], non_neg_integer(), training_params()) ->

    {ok, model()} | {error, any()}.



model_create(Features, EstimatedFeatureLength, TrainingParams) ->

    esvm_nif:model_create(Features, EstimatedFeatureLength, TrainingParams).



-spec model_load(binary()) ->

    {ok, model()} | {error, any()}.



model_load(FilePath) ->

    esvm_nif:model_load(FilePath).



-spec model_save(model(), binary()) ->

    boolean() | {error, any()}.



model_save(ModelRef, FilePath) ->

    esvm_nif:model_save(ModelRef, FilePath).



-spec model_predict(model(), features()) ->

    {ok, label()} | {error, any()}.



model_predict(ModelRef, Features) ->

    esvm_nif:model_predict(ModelRef, Features).