Academic Performance Evaluation Using Data Engineering and Multi-Model Machine Learning System

This paper aims at identifying and employing data engineering and multi-model machine learning techniques for the assessment of student's academic performance with a strong focus on data analytics approach in education. It starts with data collection and cleaning of various data types such as demographic, attendance, academic performance, and co-curricular records. These are consolidated into a unified database for training. Next, the Machine Learning model is developed and trained with the help of the chosen Random Forest. Random Forest is one of the strong ensembles learning models that build a large number of decision trees at the time of training. Training of each decision tree is carried out on the features and data samples that are part of the dataset but different from the others to improve on the degree off repetition and thereby increase the level of robustness of the final prediction. After the model training, several Python libraries are used to design a local website for collecting student details for performance prediction. The system then cross checks the input details with the trained model and provides the predicted performance output. This approach therefore ensures that the model is able to perform well across various datasets and improves its capacity to predict accurately, making it resourceful in identifying and predicting academic performance in this study.