Predicting academic performance in Educational Data Mining has been a significant research area. This involves utilizing machine learning techniques to analyze data from educational settings. Predicting student academic performance is a complex task due to the influence of multiple factors. This research uses supervised machine-learning approaches to predict students' grades and marks. The dataset used in this study is obtained from Rajya Shiksha Kendra (RSK) in Bhopal, Madhya Pradesh, India. RSK consists of four blocks: "Phanda-Rural," "Phanda-Urban," "Phanda-Old City," and "Berasia." The total number of schools in all four blocks of the Bhopal district is 3,201. The system's abundant data requires proper analysis to extract the most valuable information for planning and future development. Predicting grades and marks based on students' historical educational records is practical in assessing schools' academic performance in the Bhopal district. It serves as a valuable source of information that can be utilized in various ways to enhance the quality of education nationwide. This paper aims to assess the quality of teaching and academic performance of different schools in the Bhopal district of Madhya Pradesh, India. The acquired UDISE dataset is pre-processed in the proposed approach to ensure data quality. Genetic algorithms, decision tree classifiers, and various machine learning models were used on the dataset using student's and schools labeled academic historical data. The proposed model predicts academic performance, while the classification system predicts assessment grades. The obtained results are then analyzed. The findings demonstrate the efficiency and relevance of machine learning technology in predicting academic performance. The study also suggests the critical inclusion of various indicators/variables which help predict schools' academic performance.