This repository contains the practical exercises and solutions for building and analyzing regression models.
- Linear Regression: Build a model to explain the target variable Y using explanatory variables X1, X2, ..., Xm using a linear hypothesis function.
- Simple Linear Regression: Use gradient descent to minimize the Mean Square Error (MSE) and predict land prices from land area.
- Multi Linear Regression: Apply linear regression techniques to the Boston Housing Dataset to predict median value of owner-occupied homes (MEDV).
- Objective: Build a model with a hypothesis to predict the land price from an input land area. Minimize the cost function (MSE) using Gradient Descent.
- Dataset:
data_1.1.csv - Tasks:
- Update parameter values using Gradient Descent.
- Calculate gradients and loss function.
- Display graphical representation of the loss function.
- Test various learning rates and plot the model performance.
- Objective: Predict the values of MEDV using machine learning techniques.
- Tasks:
- Load and preprocess the dataset.
- Check for duplicate entries and handle missing values.
- Encode categorical variables and calculate the correlation matrix.
- Visualize relationships using scatter plots.
- Split the data into training and testing sets (80/20 ratio).
- Objective: Convert house price data into polynomial form and perform regression.
- Tasks:
- Apply polynomial transformations to features.
- Implement gradient descent to find the cost for normal and quadratic values.
- Objective: Implement logistic regression from scratch using gradient descent to predict binary outcomes.
- Tasks:
- Initialize the logistic regression model.
- Fit the model to the training data.
- Predict the binary class for input data.
- Use the sigmoid function to map model outputs to probabilities.
see the Practical session.docx
Feel free to fork this repository, submit pull requests, or send us suggestions on how to improve the models or analyses.