# Regression and Prediction

Everyone
• 12 videos | 1h 1m 18s
• Includes Assessment
Learn the basics of regression for prediction and inferential purposes. Also, gain an understanding of modern linear and non-linear regression for prediction and inferential purposes. Finally, get practical experience of using classical and modern regression methods for prediciton and inferential purposes.

## WHAT YOU WILL LEARN

• Gain a basic understanding of what will happen in the course Know how to use linear regression for prediction Understand anova in the sample and in the population and know how to assess the output of sample predictive performance Be able to describe why beta 1 is the regression coefficient and know why this is useful for understanding the regression coefficient Learn about nonlinear regression forms. also learn about regressions that result from replacing the squared loss function by other loss functions. Know when to use high-dimensional regressors in prediction
• Explain approximate sparsity and lasso Know how to estimate the target regression estimate in the high dimensional regression problem Be able to use cross validation Understand tree based prediction rules and ways to improve them by bagging or boosting Know what neural networks are and how they work Know when causality can and cannot be established from regression

## IN THIS COURSE

• Meet the instructor and learn about what will be taught in this course.
• Define linear regression in the population and in the sample. Learn about the best linear predictor.
• 3.  Assessment Of Prediction Quality
Learn how to understand the analysis of variance, also known as ANOVA. Also, learn how to assess the output of sample predictive performance of the sample linear regression
• 4.  Inference For Linear Regression
Learn the answer to the inference question posited in previous videos.
• 5.  Other Types Of Regression
Understand regressions using nonlinear forms as well as regressions that result from replacing the squared loss function by other loss functions
• 6.  Modern Linear Regression For High-Dimensional Data
Learn the two motivations for using high-dmensional regressors in prediction.
• 7.  High-Dimensional Sparse Models And Lasso
Learn about high-dimensional sparse models and a penalized regression method called the Lasso.
• 8.  Inference With Lasso
Find out how to use loss to answer the inference question
• 9.  Other Penalized Regression Methods: Cross-Validation
Learn other penalized regression methods. Also learn what cross-validation is.
• 10.  Trees, Random Forests, and Boosted Trees
Learn how to use trees as a predictor for data.
• 11.  Neural Networks
Find out what a neural network is.
• 12.  Causality: Can It Be Established From Regression?
Learn about causality. Find out when you can and cannot establish it from regression.