E-WASTE MANAGEMENT THROUGH DEEP LEARNING: A SEQUENTIAL NEURAL NETWORK APPROACH

Abstract

The goal of this research is to improve the management of electronic trash (e-waste) by using a Sequential Neural Network (SNN) with TensorFlow and Keras as part of an advanced deep learning technique. In order to address the growing problem of e-waste, the research collects a large amount of data from images of e-waste and then carefully preprocesses and augments those images. With precision, recall, and F1 scores of 87%, 86%, and 86%, respectively, the SNN architecture—which incorporates dropout, pooling, and convolutional layers—achieved an amazing 100% classification accuracy. These outstanding outcomes show how well the model can classify e-waste components, suggesting that it has the potential to be used in real-world scenarios. The results indicate that the SNN-based approach greatly improves the accuracy and efficiency of e-waste sorting, promoting environmental sustainability and resource conservation. By automating the sorting process, the suggested system decreases the need for manual labor, minimizes human error, and speeds up processing. The study emphasizes the model's suitability for integration into current e-waste management workflows, providing a scalable and dependable way to expedite the recycling process. Additionally, the model's real-time applicability highlights its potential to revolutionize current e-waste management practices, making a positive ecological impact. . Future research endeavors will center on broadening the dataset to include a wider range of e-waste image categories, investigating more advanced deep learning architectures, and incorporating the system with Internet of Things (IoT) devices to improve real-time monitoring and management.