Multiple Regression with Minitab

Explore multiple regression by using several predictor variables to model a response variable, using beta-coefficient equations and examples like hours studied and prior test scores.

Demonstrate multiple regression in Minitab by modeling test score from hours studied and previous test, and visualize results with scatterplots and contour and surface plots.

Analyze a two-variable multiple regression in Minitab, using scatterplots to link test scores to hours studied and prior scores, and evaluate p-values, R-squared, adjusted R-squared, and residual diagnostics.

Analyzes a two-predictor multiple regression with hours studied and previous test score. Discusses p-values, VIF, multicollinearity, R-squared, R-square adjusted, R-square predicted, ANOVA, residual plots, and surface and contour visualizations.

Demonstrate a two-predictor multiple regression in Minitab, using hours studied and gender to predict test scores, with scatterplots, regression lines, and single equation interpretation.

Analyze how hours studied and gender affect test scores using separate and single regression equations, compare R-squared and p-values, and illustrate gender differences with regression and factorial plots.

Learn to distinguish underfitting and overfitting in regression, compare straight-line and curved fits, and preview the bias-variance tradeoff for future data.

Learn the bias-variance trade-off in multiple regression with minitab, defining bias as current-data error and variance as future-data sensitivity. Understand how underfitting and overfitting relate to model complexity.

Learn how polynomial models, especially the quadratic model, balance bias and variance in non-linear relationships, and why higher degrees may not represent future data. Keep it simple with second-degree terms.

Explore fitting linear, quadratic, and cubic polynomial models to hours studied and test scores in Minitab, compare model fit, and assess performance on unseen data.

Compare linear, quadratic, and cubic regression models in Minitab for hours studied predicting test scores; assess p-values and R square values to choose the best model, with side-by-side evaluation.

Compare linear, quadratic, and cubic regression using Minitab on hours studied data, evaluate p-values, VIF, and adjusted and predicted R-square, and discover stepwise feature selection to select the best model.

Explore stepwise selection in regression with Minitab, using hours studied to predict test score, comparing linear, quadratic, and cubic terms via p-values and R square to select the best model.

Explore feature selection in multiple regression with minitab, choosing independent predictors over correlated ones, and compare best subset and stepwise selection for model reduction.

Investigate how cement components x1, x2, x3, and x4 influence heat evolved using scatterplots and multiple regression in Minitab, and introduce best subset method to choose the best model.

Compare regression models using R-square, adjusted R-square, and the standard error, then apply Mallows' Cp and the best subset method in Minitab.

Explore best subsets regression in Minitab to select the optimal two-variable model (X1 and X2) for predicting heat evolved, including quadratic terms and Mallows Cp.

Explore feature selection with stepwise methods in Minitab, comparing forward and backward elimination to the best subsets approach, using x1–x4, r-square, Mallows' Cp, and S to build a final model.

Explore stepwise feature selection in Minitab for multiple regression with x1, x2, x3, x4, including squares and interactions, comparing to best-subset and showing how stepwise yields a concise, well-fitted model.

Identify outliers in regression using a scatterplot, assess their impact on p-values and R-squared, and decide whether to remove unusual observations from the dataset.

Understand how outliers and extreme X values impact regression in Minitab, and learn how leverage, Cook's distance, and DFIT inform data cleaning decisions.

Learn how to split data into training and testing sets to evaluate regression models on unseen data, using R-squared, adjusted R-squared, and R-square predicted to gauge future performance.

Explore how to test regression models on unseen data using validation set, leave-one-out, and k-fold validation in Minitab, with 70:30 splits and practical guidance.

Explore k-fold and leave one out cross validation in Minitab, illustrating how to partition data for training and testing, perform tenfold validation, and compare model performance.

Explore building a regression model in Minitab with hours studied predicting test scores, and compare validation methods like 70:30 test split and k-fold or leave-one-out cross validation.

Learn to estimate the response from predictors using a Minitab regression model, including model validation, prediction, and interpreting heat evolved from X1, X2, X3, and X4.

Explore a medical insurance charge dataset in Minitab, import insurance.csv from Kaggle, and build a regression model to predict insurance amounts using age, BMI, children, smoker status, and region.

Apply descriptive statistics and matrix plots to explore how age, body mass index, number of children, smoking status, region, and gender relate to insurance charges.

Develop a regression model in Minitab to predict charges using age, BMI, smoking status, and categorical variables. Compare simple and limited models using p-values, R-squared, and scatterplots.

Improve the final regression model for insurance charges by adding age squared and interactions, using stepwise with validation to select predictors, achieving about 84% ten-fold R-square and reliable predictions.