Evaluation of Students' Performance in Educational Sciences and Prediction of Future Development using TensorFlow

摘要

Artificial Intelligence is the domain of computer science which includes the solving of problems in reasoning, knowledge representation, prediction, learning and perception areas. The large volume of data can be used for social media, e-learning, distance learning and e-commerce environment. Our research work includes the classification and prediction of students' performance in educational sciences. The analyzed results are forecasting the future plan in higher studies. In this work, we use TensorFlow Artificial Intelligence engine for classification. Deep learning is used for measuring academic performance in core courses such as mathematics, physics, chemistry, biology and computer Science. The performance can be measured in nonacademic activities also such as sports, yoga, art and social services. These papers gives prediction result using machine learning tools and give more comprehensive study of both academic and nonacademic activities. Here we take number of intermediate nodes from students' performance and number of deep learning objects from students' activities. The result is generated and compared using TensorFlow. The input of two thousand five hundred students' data is taken from Tamil Nadu Nagapattinam and Thirvarur Districts from education science department, 65% of data is trained data and 35% of data are test data. The accuracy factor is 75% to 85%. The prediction factor accuracy can be determined by using optimal configuration of TensorFlow engine. This result can be used for the benefit of the students to select their future studies and career development of students based on their higher secondary academic and nonacademic performance factors.