om-ashish-soni / spi-predictor

Building an AI model that predicts a student's SPI (GPA) based on their internal exam marks. The model is created using linear regression and the Gradient Descent algorithm.

SPI ( GPA ) Predictor

Introduction

• This project is a simple implementation of a linear regression model using PyTorch library to predict SPI (Standardized Precipitation Index) values based on input precipitation data.

• The dataset consists of 20 rows of input data, each with 3 features, and a corresponding SPI value.

• The model is trained using mean squared error loss and gradient descent optimization.

Requirements

The following libraries are required to run the code:

NumPy

PyTorch

Usage

• To run the code, simply execute the script in a Python environment that has the required libraries installed. The script will output the final loss, weights, and bias of the trained model.

Algorithm

• Gradient Descent Algorithm with Linear Regression is used to train this model

Code Explanation

• The input data is defined as a NumPy array with 20 rows and 3 columns. The target output is also defined as a NumPy array with 20 rows and 1 column.

• The weights and bias of the neural network model are initialized randomly and with requires_grad set to True, indicating that they need to be updated during training.

• The model is trained for a fixed number of iterations using gradient descent optimization. The loss function used is mean squared error loss.

• After training, the final loss, weights, and bias are printed to the console.

Future Improvements

• Some potential areas of improvement for this project include:

• Use a larger and more diverse dataset to improve the accuracy of the model.

• Experiment with different neural network architectures and hyperparameters to find the best model configuration.

• Implement techniques such as early stopping and regularization to prevent overfitting.