False Positives for every classification threshold.Īn ROC curve that hugs the upper left corner represents a “good” classifier, that does a fine job of separating the classes, while a curve that fallse near the line y =x represents a classifier that is not much better than guessing. The purpose of ROC (Receiver Operating Characteristic) Curves is to examine the performance of a binary classifier, by creating a graph of the True Positives vs. Once you test a data-set with a classifier, particularly a self-made classifier, it is important to examine the outcome. Weighted Averages of Scores (Model Evaluation)

Learning Curve: Part B (Model Evaluation) Learning Curve: Part A (Model Evaluation) Generating Stratified Folds (Data Preprocessing)Ĭreating Training, Validation and Test Sets (Data Preprocessing) Random Under Sampling (Class Imbalance Problem) Precision Recall Curve (Model Evaluation) ROC Curves and AUC (A General Overview).Inverse k-fold Cross Validation (Model Evaluation) Semi-supervised Learning (Learning Techniques) Setting Class Attribute (Data Preprocessing)Ĭlassification 101 using API (Classification) Merge and Append ARFF Files (Data Preprocessing) Outliers and Extreme Values (Data Preprocessing)Īttribute Selection with Knowledge Flow Environment (Data Dimensionality) Stacking Multiple Classifiers (Classification)ĭetail Cross Validation Results using API (Model Evaluation)Ĭlassification 101 with Weka Knowledge Flow Environment (Classification) Generating Non-stratified Folds (Data Preprocessing)Ĭross Validation Error Rates (Model Evaluation) Systematic Oversampling (Class Imbalance Problem)įeature Selection with Wrapper (Data Dimensionality)įeature Selection with Filter (Data Dimensionality) Need a detailed video detailing one of the following processes in Weka? Click the hyperlinks! While I have not been able to cover every topic in Weka, I was able o find many tutorials outlining more complex concepts in Weka.

How to install weka wraper