An Ensemble Deep Learning Model for Vehicular Engine Health Prediction

In the automotive industry, maintaining optimal engine health is crucial for ensuring vehicle performance and longevity. This project introduces an ensemble deep learning approach to predict vehicular engine health using a combination of Random Forest, Decision Tree, and K-Nearest Neighbors (KNN) algorithms. Each algorithm contributes uniquely to the ensemble, leveraging their strengths in classification and regression tasks. The Random Forest model achieves an accuracy of 84%, harnessing its ability to handle complex data interactions and reduce overfitting. Meanwhile, the Decision Tree model, with an accuracy of 76%, offers interpretability and captures nonlinear relationships within the data. The KNN model complements these by achieving an accuracy of 80%, leveraging instance-based learning to classify engine health based on similarity to neighboring data points. By combining these models into an ensemble, the project enhances prediction robustness and reliability. This approach not only improves accuracy but also provides a comprehensive framework for real-time engine health monitoring and proactive maintenance scheduling. The results underscore the effectiveness of ensemble deep learning in addressing critical challenges in automotive diagnostics and maintenance.