| # Implementation Plan - Student Marks Prediction using RNN | |
| This document outlines the steps to build an end-to-end Recurrent Neural Network (RNN) model to predict student marks based on the number of courses and study time. | |
| ## 1. Data Exploration & Preprocessing | |
| - Load `Student_Marks.csv`. | |
| - Inspect data quality and statistics. | |
| - Normalize features (`number_courses`, `time_study`) and target (`Marks`) using `MinMaxScaler` or `StandardScaler`. | |
| - Split the dataset into training (80%) and testing (20%) sets. | |
| - **RNN Reshaping**: Reshape the input data to `(samples, time_steps, features)`. Since this is a simple tabular dataset, we will use `time_steps = 1`. | |
| ## 2. Model Architecture | |
| - **Input Layer**: Shape `(1, 2)`. | |
| - **RNN Layer**: Use `SimpleRNN` or `LSTM` with 64 units. | |
| - **Dense Layer**: Hidden layer with 32 units, ReLU activation. | |
| - **Output Layer**: Single neuron for regression (predicted Marks). | |
| - **Compile**: Use `Adam` optimizer and `Mean Squared Error` (MSE) loss. | |
| ## 3. Training | |
| - Train for 100 epochs (adjustable). | |
| - Use a validation split to monitor overfitting. | |
| ## 4. Evaluation & Visualization | |
| - Evaluate the model on the test set. | |
| - Plot training and validation loss curves. | |
| - Compare predicted values with actual values. | |
| ## 5. Inference | |
| - Create a script to make predictions on new data. | |