Machine Learning for Classification Problems

Our goal is to provide a clear and accessible “behind the scenes” understanding of many machine learning procedures for classification that are more frequently being used in the health and behavioral sciences. We strive to strike a balance between core concepts of the algorithmic procedures along with their application and interpretation when implemented with real empirical data. This workshop is designed to provide participants with the knowledge to discern which machine learning models are appropriate for their research question and the subsequent tools to analyze their data in ways not previously possible.

This workshop is designed for graduate students, post-doctoral fellows, faculty, and research scientists from the behavioral, social, and health sciences.

Although it is recommended that participants have a working knowledge of the multiple regression model, these core concepts will be reviewed at the beginning of the workshop. Participants who seek a more in depth review prior to attending may wish to watch our Linear Regression Playlist on YouTube.

Chapter 1: Introduction
1.1 Orientation of the Workshop
1.2 Goals of Supervised Classification
1.3 History and Evolution of Classification
1.4 Relationship to Predictive Modeling

Chapter 2: Over-Fitting and Model Tuning
2.1 Evaluating Classification Models
2.2 The Problem of Over-Fitting
2.3 Model Tuning and Data Splitting
2.4 Resampling Techniques
2.5 Choosing Between Models

Chapter 3: Linear Classification Models
3.1 Logistic Regression
3.2 Linear Discriminant Analysis and Variants
3.3 Nearest Shrunken Centroids

Chapter 4: Nonlinear Classification Models
4.1 Nonlinear Discriminant Analysis
4.2 Flexible Discriminant Analysis
4.3 Support vector machines

Chapter 5: Classification Trees
5.1 Classification Tree Models
5.2 Model Aggregation: Bagging/Boosting
5.3 Random Forests
5.4 C5.0 Algorithm

Chapter 6: Additional Considerations
6.1 Addressing Class Imbalance
6.2 Variable (Feature) Importance and Selection
6.3 Type III Errors (rejecting the null for the wrong reasons)
6.4 Measurement Error
6.5 Trusting Your Model

Chapter 7: Next Steps
7.1 Classification through Self-Organized Maps and Neural Networks
7.2 Using Deep Learning for Classification
7.3 Semi-supervised Classification
7.4 Zero-shot Classification in Large Language Models

Recorded software demonstrations for machine learning are provided for R and Python. Note that both R and Python can be downloaded for free (R download; Python download).

While it is helpful to have some familiarity with R or Python, it is not necessary. The lectures which constitute the majority of the workshop are software-independent.

Participants receive lifetime access to all course materials, including complete course notes, lecture recordings, software demonstration notes, software demonstration recordings, and data and code for all examples. You can revisit these materials any time you like, without worrying about expiration dates. Pretty awesome, huh?

Content for this class is evergreen, meaning participants will always have access to the most current version of the course materials. Thus, if we replace a lecture recording or software demonstration, or update a syntax file, you will have immediate access to the updated content without needing to re-enroll in the class. Even more awesome.