A Classifying Gender Crimes with AdaBoost and Back Propagation Algorithms
Dileep Kumar Kadali, R. N. V. Jagan Mohan, M. Chandra Naik · 2024
In today’s culture, images and videos are crucial for effective work and security surveillance. By studying CCTV data, a prediction algorithm may ascertain a person’s age, gender, location, and sexual orientation. This technology can make the world safer by enabling the identification of runaways. The technology, integrated into security cameras within a mile, can screen suspects, such as a fugitive who stole millions from a typical bank. The paper explores the development of technologies that can determine a person’s age, face image, gender, and location through cameras, pictures, or videos Using Deep Learning techniques. A machine learning algorithm AdaBoost is for binary classification tasks, combining weak classifier predictions to create a strong classifier that performs well on the given data. The research focuses on data parallelism and model parallelism as two methods for distributing backpropagation computation among GPUs or nodes using Distributed Back Propagation. The study compares AdaBoost and Back Propagation in gender crime classification. The study investigates technologies like t-SNE, PCA, and ICA in daily life and their potential applications in criminal face identification, emphasizing the need for precise experimental results.