The Complete Supervised Machine Learning Models in Python

Examine the simple linear regression intuition alongside the mean model for predicting land prices from past data, using a scatter plot, mean line, and residuals to compare model performance.

Explore the mathematics behind simple linear regression, including calculating the slope and intercept, plotting scatter data, centroids, residuals, and r-squared to understand model fit.

Train your first Python machine learning model with simple linear regression, using a linear regression class, splitting data into training and testing sets, and fitting area versus price.

Test a simple linear regression model with unseen data, predicting land prices from area values using the predict function and evaluating predictions against real values for training and testing data.

Plot the regression line on testing and training data to compare predicted values with actual data, showing the model fits well.

Implement r-squared in Python for simple linear regression, using r2_score to evaluate train and test sets, and interpret 85% and 98% variation for land price predictions.

Build a simple linear regression template with sklearn to predict land price from area. Reshape the area into a two-dimensional array for the predict function.

This extra video demonstrates building a graphical template for a simple linear regression model that predicts land price from area, via a window with input and a predict button.

Explore multiple linear regression, where one dependent variable depends on many independent variables. Learn the equation with intercept and coefficients, and how to interpret effects while selecting best predictors.

Master auditing and underfitting concepts, identify overfitting risks, and use model diagnosis to improve linear regression; learn how correlation reveals relationships and prediction reliability.

Investigate null and alternative hypotheses, interpret p-values, and use R-squared and adjusted R-squared to guard against noncontributing variables in multiple linear regression.

Diagnose model to improve reliability and simplicity by removing nonsense variables and ensuring predictors relate to the dependent variable, avoid multicore linearity by keeping one predictor (distance, area, crime rate).

Explore how dummy variables convert categorical data into numerical representation for multiple linear regression, using city examples like Zurich and Geneva and expanding to three categories with additional columns.

Explore forward selection and backward elimination as feature selection methods in regression. Apply p-value and adjusted r-squared criteria across stepwise procedures to identify impactful variables.

Implement multiple linear regression in Python with scikit-learn, using area, distance, and crime rate to predict price, then split data into training and testing sets and fit the model.

Use Python multiple linear regression to predict land price from area, distance, and crime rate; compare predicted and actual values and explore variable selection to manage multicollinearity.

Evaluate a multiple linear regression model in python by comparing r-squared across area, distance, and crime rate. Address multicollinearity by removing redundant features and ensuring proper data reshaping.

Build and test a multiple linear regression template to predict land price from area, and then from area, distance, and crime rate, using NumPy reshaping and a predict function.

Learn how to create and integrate dummy variables for categorical data in Python using pandas get_dummies, including dropping one category and concatenating with the dataset for regression models.

Establish a simple linear regression with area predicting price, select the area feature, perform train-test split, and visualize the regression line on a scatter plot.

Develop a multi-variable linear regression template in Python to predict land prices using area, distance, and crime rate.

Visualize ball distance data using matplotlib, plot a scatter of angle versus distance, and compare linear and polynomial regression lines to assess fit.

Train a linear regression model on the data, convert inputs to two-dimensional arrays, plot the regression line, evaluate with the R2 score to gauge underfitting, and preview polynomial regression.

Explore polynomial regression by expanding features with degree two into polynomial features, fitting a linear model, and evaluating with a regression score, achieving about 0.99 on the example.

Explore bias and variance in regression and how regularization through ridge regression and lasso shrinks coefficients to curb overfitting and underfitting, while elastic net combines penalties.

Explore ridge regression intuition, a penalty-based method that reduces model complexity and variance by shrinking coefficients through the loss function with lambda, improving generalization.

Explore ridge regression and simple linear regression in Python, prepare height-weight data, train and test with two-dimensional inputs, evaluate with r-squared, and visualize the regression fit.

Explore ridge regression by adding a small bias with alpha, train on data, and compare the training and testing squared error to control overfitting.

Build a ridge regression template in Python to predict weight from height. Transform height to two-dimensional input with NumPy reshape or an array, then predict weights.

Build a ridge regression template in Python that predicts weight from height using a user input and a live display, adapting a simple linear regression template.

Explore lasso regression intuition and how it uses a penalty in regularization to shrink coefficients, compare it to ridge regression, reduce overfitting, and enable feature selection for simpler models.