Multi-Class Classification of Vector Borne Diseases using Convolution Neural Network
Darshanaben Dipakkumar Pandya, Shailesh Kantilal Patel, Abdul Hamid Qureshi, Akashgiri Jashvantgiri Goswami, Sheshang Degadwala, Dhairya Vyas
Abstract
The worldwide burden of illness and death caused by vector-borne pathogens is substantial. The capacity to properly categorize these illnesses is crucial for the development of efficient preventative and therapeutic measures. A multi-class classification model for vector-borne illnesses is proposed in this research work. This model makes use of Convolutional Neural Networks (CNN). To train and test the model, a publicly accessible data set including information on mosquito species and the illnesses they transmit are utilized. Proposed findings show that the suggested model has an overall accuracy of 98% in classifying vector-borne illnesses. Extensive trials are processed to see how changing the model's hyperparameters affected its efficiency. Proposed results indicate that model performance is sensitive to the parameters with which the model is configured, including the number of filters, kernel size, and pooling size. The suggested multi-class classification approach can accurately categorize vector-borne illnesses, which is useful for identifying patterns and ultimately creating more targeted preventative and therapeutic measures.