Merge pull request #150 from sanity/addLogistic

fixed LogisticRegressionBuilderBug relating to redundant DataTransfor…

pom.xml

@@ -32,7 +32,7 @@
         be accompanied by a bump in version number, regardless of how minor the change.
    0.10.1 -->
 
-    <version> 0.10.7</version>
+    <version> 0.10.9</version>
 
     <repositories>
         <repository>

src/main/java/quickml/supervised/classifier/logisticRegression/LogisticRegressionBuilder.java

@@ -32,12 +32,11 @@ public class LogisticRegressionBuilder<D extends LogisticRegressionDTO<D>> imple
     public StandardDataTransformer<D> logisticRegressionDataTransformer;
 
     private ProductFeatureAppender<ClassifierInstance> productFeatureAppender;
-    private DataTransformer<ClassifierInstance, SparseClassifierInstance, D> dataTransformer;
     GradientDescent<SparseClassifierInstance> gradientDescent = new SparseSGD();
     private int minWeightForPavBuckets =2;
 
     public LogisticRegressionBuilder(StandardDataTransformer<D> dataTransformer) {
-        this.dataTransformer = dataTransformer;
+        this.logisticRegressionDataTransformer = dataTransformer;
     }
 
     public LogisticRegressionBuilder<D> productFeatureAppender(ProductFeatureAppender<ClassifierInstance> productFeatureAppender) {
@@ -67,7 +66,7 @@ public LogisticRegressionBuilder<D> poolAdjacentViolatorsMinWeight(int minWeight
 
     @Override
     public D transformData(List<ClassifierInstance> rawInstances){
-            return dataTransformer.transformData(rawInstances);
+            return logisticRegressionDataTransformer.transformData(rawInstances);
     }
 
 

src/main/java/quickml/supervised/classifier/logisticRegression/StandardDataTransformer.java

@@ -18,12 +18,6 @@
  * Created by alexanderhawk on 10/14/15.
  */
 public abstract class StandardDataTransformer<D extends LogisticRegressionDTO<D>> implements DataTransformer<ClassifierInstance, SparseClassifierInstance, D> {
-    //to do: get label to digit Map and stick in DTO (and transform to logistic regression eventually)
-    //make LogisticRegressionBuilder use this class and not be tightly coupled to mean normalization (e.g. allow log^2 values)
-    //make cross validator take a datetransformer (specifically, the Logistic regression PMB, and then do the data normalization
-    // and set the date time extractor)
-
-
     /**
      * class provides the method: transformInstances, to convert a set of classifier instances into instances that can be processed by
      * the LogisticRegressionBuilder.
@@ -32,9 +26,7 @@ public abstract class StandardDataTransformer<D extends LogisticRegressionDTO<D>
      * product feature appendation as well as common co-occurences should be hyper-params within logistic regression.
      *
      */
-    /*Options, wrap logistic regression? in a new logistic regression class that has a logistic reg transformer?
-            * Or change sparse classifier instance as the the type of Logistic Regression?  I almost prefer this.  So now to use it...one just passes in a normal list of training instances
-    */
+
 
     protected ProductFeatureAppender<ClassifierInstance> productFeatureAppender;
 

src/test/java/quickml/supervised/classifier/logRegression/LogisticRegressionBuilderTest.java

@@ -1,5 +1,7 @@
 package quickml.supervised.classifier.logRegression;
 
+import com.google.common.collect.Iterables;
+import com.google.common.collect.Maps;
 import org.junit.Ignore;
 import org.junit.Test;
 import org.slf4j.Logger;
@@ -21,9 +23,16 @@
 import quickml.supervised.crossValidation.lossfunctions.classifierLossFunctions.WeightedAUCCrossValLossFunction;
 import quickml.supervised.dataProcessing.instanceTranformer.CommonCoocurrenceProductFeatureAppender;
 import quickml.supervised.ensembles.randomForest.randomDecisionForest.RandomDecisionForestBuilder;
+import quickml.supervised.predictiveModelOptimizer.FieldValueRecommender;
+import quickml.supervised.predictiveModelOptimizer.PredictiveModelOptimizer;
+import quickml.supervised.predictiveModelOptimizer.fieldValueRecommenders.FixedOrderRecommender;
 import quickml.supervised.tree.decisionTree.DecisionTreeBuilder;
 
 import java.util.List;
+import java.util.Map;
+
+import static quickml.supervised.classifier.logisticRegression.LogisticRegressionBuilder.MIN_OBSERVATIONS_OF_ATTRIBUTE;
+import static quickml.supervised.classifier.logisticRegression.SparseSGD.*;
 
 /**
  * Created by alexanderhawk on 10/13/15.
@@ -121,4 +130,56 @@ public void testDiabetesInstances() {
         logger.info("RF out of time loss: {}", simpleCrossValidator.getLossForModel());
     }
 
+    @Ignore
+    @Test
+    public void optimizerTest(){
+
+        List<ClassifierInstance> instances = InstanceLoader.getAdvertisingInstances().subList(0,1000);
+        CommonCoocurrenceProductFeatureAppender productFeatureAppender = new CommonCoocurrenceProductFeatureAppender<>()
+                .setMinObservationsOfRawAttribute(35)
+                .setAllowCategoricalProductFeatures(false)
+                .setAllowNumericProductFeatures(false)
+                .setApproximateOverlap(true)
+                .setMinOverlap(20)
+                .setIgnoreAttributesCommonToAllInsances(true);
+
+        DatedAndMeanNormalizedLogisticRegressionDataTransformer lrdt = new DatedAndMeanNormalizedLogisticRegressionDataTransformer()
+                .minObservationsOfAttribute(35)
+                .usingProductFeatures(false)
+                .productFeatureAppender(productFeatureAppender);
+
+        LogisticRegressionBuilder<MeanNormalizedAndDatedLogisticRegressionDTO> logisticRegressionBuilder =  new LogisticRegressionBuilder<MeanNormalizedAndDatedLogisticRegressionDTO>(lrdt)
+                .calibrateWithPoolAdjacentViolators(false)
+                .gradientDescent(new SparseSGD()
+                                .ridgeRegularizationConstant(0.1)
+                                .learningRate(.0025)
+                                .minibatchSize(1000)
+                                .minEpochs(500)
+                                .maxEpochs(500)
+                                .minPredictedProbablity(1E-3)
+                                .sparseParallelization(true)
+                );
+        double start = System.nanoTime();
+        EnhancedCrossValidator<LogisticRegression, ClassifierInstance, SparseClassifierInstance, MeanNormalizedAndDatedLogisticRegressionDTO> enhancedCrossValidator = new EnhancedCrossValidator<>(logisticRegressionBuilder,
+                new ClassifierLossChecker(new WeightedAUCCrossValLossFunction(1.0)),
+                new OutOfTimeDataFactory(0.25, 48), instances);
+
+
+
+
+
+        Map<String, FieldValueRecommender> sgdParams = Maps.newHashMap();
+        sgdParams.put(RIDGE, new FixedOrderRecommender(.0001));//;, .001, .01, .1, 1));//MonotonicConvergenceRecommender(numTreesList, 0.01));
+        sgdParams.put(MIN_EPOCHS, new FixedOrderRecommender(8000));// 16000));
+        sgdParams.put(MAX_EPOCHS, new FixedOrderRecommender(16000));//, 3200));
+        sgdParams.put(LEARNING_RATE, new FixedOrderRecommender(.0025));//, .001, .005));//11, 14, 16 //Pbest 12
+        sgdParams.put(MIN_OBSERVATIONS_OF_ATTRIBUTE, new FixedOrderRecommender(20, 50));// 16000));
+        PredictiveModelOptimizer modelOptimizer = new PredictiveModelOptimizer(sgdParams, enhancedCrossValidator, 2);
+
+
+
+
+        logger.info("Optimal sgd parameters: {}", modelOptimizer.determineOptimalConfig());
+    }
+
     }