Software development system based on logistic regression algorithm

By using Word2Vec model and logistic regression algorithm in code defect prediction, combined with the module interaction strength, a comprehensive feature vector is formed, which solves the shortcomings of traditional methods in model structure settings and improves the accuracy and reliability of code defect prediction.

CN119645809APending Publication Date: 2025-03-18SHANGTOU INFORMATION SERVICES (SHANGHAI) CO LTD
9 Cites 0 Cited by

Patent Information

Application Number
CN202411765267.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-04
Publication Date
2025-03-18

AI Technical Summary

Technical Problem

Traditional code defect prediction methods based on logistic regression are not targeted when setting the model structure, cannot effectively receive vectors containing rich semantics and interactive information, and it is difficult to mine complex code defect patterns.

Method used

The annotations are parsed through the code processing unit and converted code elements into vectors using the Word2Vec model. The integration unit divides modules based on class definition and analyzes the interaction strength between modules, forms a comprehensive feature vector, and uses logistic regression algorithm for training and prediction.

Benefits of technology

It improves the accuracy and reliability of code defect prediction, can better mine complex code defect patterns, and optimize model parameters through gradient descent algorithms to improve the interpretability and practicality of the prediction results.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of machine learning, in particular to a logistic regression algorithm-based software development system, which comprises a code processing unit, an extraction and integration unit and a logistic regression unit, the code processing unit analyzes annotations to obtain semantics and converts code elements into vectors by using a Word2Vec model, the extraction and integration unit divides modules based on class definition, and the logistic regression unit divides the modules into the logistic regression algorithm-based software development system. The analysis module calculates the interaction strength and integrates the semantic vector and the interaction strength into a comprehensive feature vector, the logic regression unit inputs the comprehensive feature vector into the model, performs training learning and code defect prediction by using an algorithm, outputs a result and compares the result with the actual result to obtain an evaluation index, and then optimizes model parameters by using a gradient descent algorithm. According to the system, code semantics and module interaction are comprehensively considered, a model structure and a training method are optimized, the code defect prediction accuracy is improved, and effective support is provided for software development quality control.
Need to check novelty before this filing date? Find Prior Art