Example usage for weka.classifiers.functions LinearRegression setRidge

Introduction

In this page you can find the example usage for weka.classifiers.functions LinearRegression setRidge.

Prototype

public void setRidge(double newRidge) 

Source Link

Document

Set the value of Ridge.

Usage

From source file:edu.utexas.cs.tactex.utils.RegressionUtils.java

License:Open Source License

public static Double leaveOneOutErrorLinRegLambda(double lambda, Instances data) {

    // MANUAL /* w w w .  j ava  2 s. c  om*/

    // create a linear regression classifier with Xy_polynorm data
    LinearRegression linreg = createLinearRegression();
    linreg.setRidge(lambda);

    double mse = 0;
    for (int i = 0; i < data.numInstances(); ++i) {
        log.info("fold " + i);
        Instances train = data.trainCV(data.numInstances(), i);
        log.info("train");
        Instances test = data.testCV(data.numInstances(), i);
        log.info("test");
        double actualY = data.instance(i).classValue();
        log.info("actualY");
        try {
            linreg.buildClassifier(train);
            log.info("buildClassifier");
        } catch (Exception e) {
            log.error("failed to build classifier in cross validation", e);
            return null;

        }
        double predictedY = 0;
        try {
            predictedY = linreg.classifyInstance(test.instance(0));
            log.info("predictedY");
        } catch (Exception e) {
            log.error("failed to classify in cross validation", e);
            return null;
        }
        double error = predictedY - actualY;
        log.info("error " + error);
        mse += error * error;
        log.info("mse " + mse);
    }
    if (data.numInstances() == 0) {
        log.error("no instances in leave-one-out data");
        return null;
    }
    mse /= data.numInstances();
    log.info("mse " + mse);
    return mse;

    //     // USING WEKA 
    // 
    //     // create evaluation object
    //     Evaluation eval = null;
    //     try {
    //       eval = new Evaluation(data);
    //     } catch (Exception e) {
    //       log.error("weka Evaluation() creation threw exception", e);      
    //       //e.printStackTrace();    
    //       return null;
    //     }
    //     
    //     // create a linear regression classifier with Xy_polynorm data
    //     LinearRegression linreg = createLinearRegression();
    //     linreg.setRidge(lambda);
    // //    try {
    // //      linreg.buildClassifier(data);
    // //    } catch (Exception e) {
    // //      log.error("FAILED: linear regression threw exception", e);
    // //      //e.printStackTrace();    
    // //      return null;
    // //    }
    //     
    //     // initialize the evaluation object
    //     Classifier classifier = linreg;
    //     int numFolds = data.numInstances();
    //     Random random = new Random(0);
    //     try {
    //       eval.crossValidateModel(classifier , data , numFolds , random);
    //     } catch (Exception e) {
    //       log.error("crossvalidation threw exception", e);
    //       //e.printStackTrace();    
    //       return null;
    //     }
    //     
    //     double mse = eval.errorRate();
    //     return mse;
}

---

From source file:edu.utexas.cs.tactex.utils.RegressionUtils.java

License:Open Source License

public static WekaLinRegData createWekaLinRegData(int timeslot, Instances X, Double[] yvals,
        ArrayList<Double> candidateLambdas) throws Exception {
    WekaLinRegData result;/*  www . ja  v  a 2 s .  com*/

    // normalize
    Standardize standardize = new Standardize();
    try {
        standardize.setInputFormat(X);
    } catch (Exception e) {
        log.error("PolyRegCust.predictNumSubs() data standardizing exception", e);
        throw e;
    }
    Instances nrmFeatures = RegressionUtils.featureNormalize(X, standardize);
    log.info("normalized features " + nrmFeatures);

    // add y to X since this is what weka expects
    Instances Xy = RegressionUtils.addYforWeka(nrmFeatures, yvals);

    // run cross validation for lambda
    Double bestLambda = findBestRegularizationParameter(Xy, candidateLambdas);
    if (null == bestLambda) {
        String message = "best regularization parameter is null, cannot predict";
        log.error(message);
        throw new Exception(message);
    }

    // run linear regression
    LinearRegression linearRegression = RegressionUtils.createLinearRegression();
    linearRegression.setRidge(bestLambda);
    try {
        linearRegression.buildClassifier(Xy);
        log.info("theta " + Arrays.toString(linearRegression.coefficients()));
    } catch (Exception e) {
        log.error("PolyRegCust.predictNumSubs() buildClassifier exception", e);
        throw e;
    }

    result = new WekaLinRegData(standardize, linearRegression, timeslot);
    return result;
}

---