The Implementation of Neural Network On Determining the Determinant Factors Towards Students’ Stress Resistance

Abstract

Stress is a condition that commonly felt by almost everyone, including college student. Naturally, human beings have a stress resistance in various levels. On previous research, an artificial neural network with backpropagation algorithm has been built to predict stress resistance level among college student. The level of stress resistance was predicted using four determinant factors i.e. frustration tolerance, conflict tolerance, anxiety tolerance, and tolerance to perceive changes as a challenge. On that research, the artificial neural network can predict stress resistance among college student correctly with an accuracy reach 75% after being trained up to 10334 epochs. On this research, dimensional reduction method will be applied on the determinant factors of stress resistance to eliminate disturbance factor and increase the accuracy of artificial neural networks in predicting stress resistance among college student. After the network was trained without disturbance factor i.e. anxiety tolerance, better network obtained. Experimental result showed that artificial neural network not only has better accuracy up to 81.5% but also faster training process which is only take 5000 epochs. Based on these results, the determinant factors of stress resistance among college student are: frustration tolerance, conflict tolerance, and tolerance to perceive changes as a challenge.