Student-Performance-Predictor

Linear Regression

Aim

• To Develop a secondary school student performance prediction tool to identify various variables that affect educational success and failure in secondary education based on real-world data on two core subjects - mathematics and Portuguese, which provide fundamental knowledge for success in the remaining subjects - collected using school reports and questionnaires.Techstack: Python 3.7, numpy, pandas, matplotlib, sklearn, seaborn.

Results

• I Developed a Linear regression supervised machine learning classification model whose numerical predicted value output ranges from I to V: I-(excellent/very good), II-(good), III-(satisfactory), IV-(sufficient) and V-(fail) based on the Erasmus grade conversion system.
• Math class had a 0.03548 variance score and 1.887629 root means squared error (RMSE).
• Portuguese class had a 0.09151 variance score and 1.803660 root mean squared error (RMSE).

Data

❖ Non-pre-processed dataset sourced from: https://archive.ics.uci.edu/dataset/320/student+performance

❖ There are 33 features/attributes before data preprocessing for both Mathematics and Portuguese classes. The Mathematics class has 395 records while the Portuguese class has 649 records.

Work Flow

STEPS 1

• Import libraries

STEP 2

• Fetch data and load both mathematics and Portuguese CSV files into separate pandas data frames

STEP 3

• Data Wrangling: Transform and Analyse the data.

STEP 4

• Determine the Target variable and create an Explanatory variable

• Testing and Training data split

• Build the Linear Regression model

• Run Predictions

STEP 5

• Evaluate the Model

• Visualise results