| _version_ | 1866902274145517568 |
|---|---|
| author | DIVIDEVARA SAI TEJA , G. PRAVEEN BABU |
| author_facet | DIVIDEVARA SAI TEJA , G. PRAVEEN BABU |
| contents | <p>The rising complexity of modern software has made defect management a persistent challenge, as traditional<br>testing and manual reviews often miss fault-prone modules. These overlooked defects increase development costs,<br>reduce system reliability, and can even lead to project failure. To address this, IntelliGuard introduces a machinelearning-driven solution that predicts defective modules before they cause downstream issues. By leveraging<br>historical defect data and extracting meaningful metrics, the system enables teams to focus on high-risk<br>components and optimize testing efforts. At its core, IntelliGuard uses a structured ML pipeline with<br>preprocessing, feature selection, and a Random Forest classifier designed to handle imbalanced datasets through<br>weighted learning.IntelliGuard is deployed as an interactive Streamlit application that supports both singlemodule analysis and bulk prediction via CSV uploads. The app also offers real-time visualizations such as<br>performance metrics, confusion matrix plots, precision–recall curves, and feature-importance charts. Extensive<br>experiments show that the system effectively balances precision and recall, ensuring reliable detection of both<br>defective and clean modules.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17954578 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | INTELLIGUARD: ENHANCED SOFTWARE DEFECT FORECASTING WITH ML DIVIDEVARA SAI TEJA , G. PRAVEEN BABU Software Defect Prediction, Machine Learning, Quality Assurance, RandomForestClassifier, Static Code Metrics, Data-Driven Testing, Proactive Analysis, Streamlit, Scikit-learn. <p>The rising complexity of modern software has made defect management a persistent challenge, as traditional<br>testing and manual reviews often miss fault-prone modules. These overlooked defects increase development costs,<br>reduce system reliability, and can even lead to project failure. To address this, IntelliGuard introduces a machinelearning-driven solution that predicts defective modules before they cause downstream issues. By leveraging<br>historical defect data and extracting meaningful metrics, the system enables teams to focus on high-risk<br>components and optimize testing efforts. At its core, IntelliGuard uses a structured ML pipeline with<br>preprocessing, feature selection, and a Random Forest classifier designed to handle imbalanced datasets through<br>weighted learning.IntelliGuard is deployed as an interactive Streamlit application that supports both singlemodule analysis and bulk prediction via CSV uploads. The app also offers real-time visualizations such as<br>performance metrics, confusion matrix plots, precision–recall curves, and feature-importance charts. Extensive<br>experiments show that the system effectively balances precision and recall, ensuring reliable detection of both<br>defective and clean modules.</p> |
| title | INTELLIGUARD: ENHANCED SOFTWARE DEFECT FORECASTING WITH ML |
| topic | Software Defect Prediction, Machine Learning, Quality Assurance, RandomForestClassifier, Static Code Metrics, Data-Driven Testing, Proactive Analysis, Streamlit, Scikit-learn. |
| url | https://doi.org/10.5281/zenodo.17954578 |