USING FEATURE SELECTION AND ACO ALGORITHM FOR OPTIMIZING SMART CLASSROOM

Abstract

The smart education had a huge impact on learning and teaching, so it must be effective and highly efficient. An efficient smart campus or smart classroom will make the learning more and more easily, the students could learn and give the best activities. In addition, the teachers will be able to make right decisions. To achieve this goal, the smart classroom's conditions must be ideal. Since ACO (ant colony optimization algorithm) is a meta heuristic algorithm, in this paper, it is found that ACO, in conjunction with a machine learning classifier, was an effective method used in feature selection for selecting best features from an intelligent campus data set to create an environment that is conducive to academic success and student learning, such as (humidity and temperature), lighting and sound pressure levels, wind direction, and raw rainfall amounts (among other variables). In this contribution to get the most accurate results, the ACO algorithm was combined with a logistic regression classifier that was used to select the best features. The accuracy of the proposed model was 0.927438624 and 0.898268071 for two sets of data back to the School of Design and Environment 4, building located at the National University of Singapore