Intelligent Monitoring and Physical Training for College Students Based on Principal Component Analysis with Feed Forward Neural Network

In recent days, the intelligent monitoring and physical training for college students involved in analysing of physical activities and health monitoring. It explores the benefits of regular exercise and the potential of Artificial Intelligence (AI)- driven systems in order to promote physical well-being. The existing model's performance is compromised by variations in lighting and the individual body types which affected its accuracy in real-world scenarios. In this research, Principal Component Analysis with Feed Forward Neural Network (PCA-FFNN) is proposed to learn complex pattern by reducing dimensionality for improved accuracy. Initially, the data is collected from Education and Training Records (ETR) dataset and then undergoes preprocessing with min-max normalization to normalize data into a particular range. After that, the PCA is incorporated to extract relevant features by reducing its dimensionality where the features are extracted in a high dimensional space as principal components. In addition, the PCA identifies key indicators and FFNN evaluates physical training of students. From the results, the proposed PCA-FFNN attained better results in terms of accuracy (95%), precision (94%), recall (92%) and Fl-score (93%) when compared to existing PCA-NN.