student performance predictor pbl activity.pptx

Dd

1. Manual IMPLENTATION MACHINE LEARNING MODEL FOR Student performance predictor Student performance predictor

2. Introduction and Objective Predictive Need: To utilize student academic and demographic data to accurately predict their final performance or score (Pass/Fail). Goal: Create a model that can identify students at risk early, allowing for timely educational intervention. Benefit: A cost-effective, interpretable, and efficient solution to improve educational outcomes. Objective: To manually implement and evaluate a machine learning model capable of machine learning model capable of predicting a students final academic performance based on demographic and historical Features.

3. M achine Learning Algorithms used: Logistic Regression(Base line linear classification). Decision Tree Classifier(non linear classification). Random Forest Classifier(ensemble learning using multiple disese prediction.)

4. Data set information Data Set Name : Student performance factors dataset Source : UCI ML Repository Description : Dataset provides comprehensive overview of various factors affect student performance in exams. Features(columns): 20 Common features : age, gender, physical activity, school type, family income, motivation level, attendance, hours studied, previous scores, parent involvement . Target:0=Low/Fail performance,1=High/pass perfromance

5. I mplementation :

6. k +=1

7. Conclusion Model successfully provides an interpretable and efficient measure of student performance. Three machine learning models were trained and evaluated: Logistic Regression: Highest accuracy – 90% accuracy Decision Tree Classifier: 85% accuracy Random Forest Classifier: 89% accuracy The Logistic Regression model performed the best, indicating that a linear relationship between academic/demographic parameters and final grade risk is highly predictive for this dataset/ Overall, this model provides a cost-effective, interpretable, and efficient solution for predicting Students performance and machine learning can be powerful tool to assist educatorsn in early detection of academic struggle, imporving support allocation and student success.

8. THANKYOU! Presented by: G. Ganga 23EG107E18 S. Sai Santhoshini 23EG107E56 V. Vasudha 23EG107E63 V. Reetu 23EG107E64