Smart Bin: Smart Waste Segregation System Using Machine Learning and Convolutional Neural Network

Abstract

The system utilized the Rapid Application Development Methodology for the development of the system. The model that was embedded in the system was trained and tested using Google Colab IDE, the proponents also performed split testing on the model to assess its performance with the new data and to produce unbiased results. Moreover, validation metrics such as Recall, Precision, F1-Score, and Cohen Kappa were conducted to assess the accuracy of the model. Developmental and descriptive methods were employed as the proposed study’s research design. The respondents of the study include 50 Information Technology students at Sta. Teresa College; the respondents were given assessment sheets to evaluate the performance of the developed system based on the system’s Functionality, Reliability, Portability and Maintainability. The statistical treatments used in the study were weighted mean, t-test, and standard deviation.