APPARATUS AND METHOD FOR CLASSIFICATION OF TRUE AND FALSE POSITIVIES OF WEAPON SYSTEM SOFTWARE STATIC TESTING BASED ON MACHINE LEARNING

摘要

The present invention provides a false alarm product database in which product templates for each type of software static test violation are recorded and stored; A data integration unit for merging data extracted from a plurality of sources; A data preprocessing unit that performs predetermined data preprocessing to input the merged data through the data integration unit into the machine learning algorithm; By separating the preprocessed integrated data into a train set and a test set, cross validation is performed, and learning evaluation is performed on a plurality of machine learning algorithms and hyperparameters to optimize machine learning algorithms and hyperparameters. A model selection unit to select; A data prediction unit that receives the preprocessed integrated data and applies the optimal machine learning algorithm and hyperparameters selected by the optimal model selection unit to determine whether the static test violations are true or false; And a false alarm output creation unit that generates a false alarm report by referring to the template information of the false alarm output database for false alarms determined to be false by the data prediction unit. It relates to an apparatus and method capable of automating alarm classification.