REVOLUTIONIZING DNA SPECIES CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK

Abstract

Precise DNA species classification is essential in multiple domains, including forensics, medical diagnostics, and archaeology. Conventional classification methods, including sequence alignment techniques, frequently experience significant computational expenses and restricted scalability. This paper introduces a method for DNA species classification utilizing a Convolutional Neural Network (CNN), achieving an outstanding accuracy of 98.4% on established genomic datasets. Our model utilizes CNN's capacity to autonomously extract essential sequence features, facilitating accurate and efficient species identification. DNA sequences undergo preprocessing through one-hot encoding and k-mer embeddings to enhance feature representation. The experimental findings indicate that our CNN model substantially surpasses traditional methods in terms of accuracy and computational efficiency. This method's superior classification performance could transform DNA analysis in forensic investigations, medical research, and archaeological studies, offering a rapid and dependable solution for species identification.