Q-Learning Embedded Sine Cosine Algorithm for Optimizing Deep CNN in Dengue Disease Spread Detection

In tropical and subtropical areas, dengue fever, which is brought on by the dengue virus and spread by Aedes mosquitoes, continues to pose a serious threat to public health. It is essential to identify dengue infection early on to care for patients effectively and stop the virus from spreading. Even if they are accurate, traditional diagnostic techniques like polymerase chain reaction (PCR) and serological testing are time-intensive. In recent days, deep learning (DL) and Deep Convolutional Neural Networks (DCNN) have shown potential in medical diagnosis, it can automatically extract and learn hierarchical features from complex data. This paper presents an innovative framework for enhancing dengue detection using DCNN. The proposed methodology incorporates Robust Scaling (RS) for preprocessing, a Genetic Algorithm with Ensemble Feature Ranking (GA-eFR) for feature extraction, and a Q-Learning Embedded Sine Cosine Algorithm (QLE-SCA) for optimizing the DCNN. The model was tested on the OpenDengue dataset, demonstrating significant improvements in performance metrics. From the results obtained, the proposed RS + GA-eFR + QLE-SCA + DCNN produces accuracy of 94.40%, precision of 0.92, recall of 0.90 and F1-score of 0.94, respectively. The tool used is Jupyter Notebook and the language used is python.