Artificial Intelligence #1: Linear & MultiLinear Regression

Explore the theory of linear regression and the relationship between y and X. Estimate parameters with ordinary least squares and apply simple linear regression for a single predictor X.

Learn to implement basic linear regression using random numbers as data, train a linear model with X and y, reshape inputs, and evaluate predictions with a test set.

Define random data and build a linear regression model, calculate the r score and mean square error, and visualize the regression line using train and test data.

Build a linear regression model with the diabetes dataset by loading data, selecting features, splitting into train and test sets, training the model, and examining coefficients and predictions.

Explore linear regression on a diabetes dataset by inspecting coefficients, evaluating with r-squared and mean squared error, and visualizing predictions versus real data with scatter plots and a best-fit line.

Explore the Boston housing dataset in scikit-learn to build a linear regression model. Load the data, inspect features and target, and split into train and test sets.

Build and evaluate a linear regression model for the Boston houses dataset, fitting coefficients, testing with X and y, and assessing accuracy via r-squared and mean squared error, with plots.

Build and evaluate a linear regression model on a random one-feature dataset, adding noise and using a train-test split to assess performance with R-squared and mean squared error.

Explore the theory of multilinear regression, linking y to two or more independent variables X, and learn to estimate coefficients and evaluate fit with r-squared and mean squared error.

Apply multilinear regression to predict global temperature using land max and min temperatures as features in python with pandas. Use train/test split, scaling, and metrics like mean squared error.

Learn to apply a two-input linear regression model to estimate global temperature, training on data, computing x1 and x2 coefficients, and assessing accuracy via mean squared error or r-squared.

Use a three-variable multivariate linear regression on TV, radio, and newspaper advertising to forecast sales, evaluate coefficients, and guide data-driven spending toward TV and radio.

Visualize real versus predicted data with color-coded scatter plots and 3D plots, and apply multilinear regression to advertising data to estimate sales.

Generate a synthetic dataset with five features and one target, add noise, split into train and test sets, train a multilinear regression model, and visualize predictions versus actual values.