One Class Classification By Ensembles Of Regression Models
python occer-realvalueddata.py filename.csv
- Program needs only one argument. Either a csv file or an arff file.
- In the case of csv file, no column names should be present. Last column in the csv file represents the label with values 1 for minority and 0 for majority. Same convention for labels is used for arff files.
OCCER algorithm converts a OCC (One class classification problem) into many regression problems in the original feature space such that each feature of the original feature space is used as the target variable in one of the regression problems. The errors of regression of a data point by all the regression models are used to compute the outlier score of the data point.
- python 3
- pandas
- pyod
- sklearn
- keras
- tensorflow
- liac-arff
def runColumnWiseRegressorAlgorithm()
Add a new regression algorithm of interest in above method by adding an if clause.
yet to come