MATH 257 - Data Modeling (Spring 2020)

\(~\)

Schedule and Syllabus

Welcome to course website for MATH 257 - Data Modeling! On this page you can find all materials we’ll use throughout the semester, starting with the syllabus linked below:

Click here to view the course syllabus

You can find course content by scrolling, or by using the navigation bar in the upper left. Please note that some material may not be posted until we’ve reached that point in the course.

Lecture Slides

Lectures in this class will be relatively short, consisting of roughly 20-30 minutes of presentation at the beginning of each class meeting. They will focus on the fundamental concepts and theory that are necessary precursors to the topics for that day/week.

Week 1 - Introduction (1/19 and 1/21)
- What is modeling?
- Review of statistical inference
Week 2 - Data Exploration (1/26 and 1/28)
- Confounding variables
- Data transformations
Week 3 - Basic Models - Linear Regression (2/2 and 2/4)
- Model specification
- Inference and prediction
Week 4 - Basic Models - K-nearest Neighbors (2/9 and 2/11)
- Academic holiday on Tuesday 2/9
- Model evaluation and comparison
Week 5 - Midterm Project #1 (2/16 and 2/18)
- Check-in meetings and work time on Tuesday
- Presentations on Thursday
Week 6 - Multiple Regression - Concepts (2/23 and 2/25)
- Categorical predictors
- Numerical predictors
Week 7 - Multiple Regression - Model Evaluation (3/2 and 3/4)
- Hypothesis testing
- Model selection criteria
Week 8 - Multiple Regression - Model Building (3/9 and 3/11)
- Interactions
- Influence Diagnostics
Week 9 - Multiple Regression - Case Study (3/16 and 3/18)
Week 10 - Midterm Project #2 (3/23 and 3/25)
Week 11 - Binary Outcomes - Logistic Regression (3/30 and 4/1)
- Introduction
- Statistical Inference
Week 12 - Logistic Regression - Model Summaries and Selection (4/6 and 4/8)
- Measuring Model Performance
- Model Selection Criteria
Week 13 - Alternatives to Regression (4/13 and 4/15)
- Classification and Regression Trees
- Academic holiday on Thursday 4/15
Week 14 - Alternatives to Regression (4/20 and 4/22)
- Random Forests
- Project meetings (round 1)
Week 15 - Final Project (4/27 and 4/29)
- Project meetings (round 2)

In-class Labs

The majority of in-person class time will be devoted to working on hands-on labs. These assignments will focus on connecting the concepts/theory from lectures with practical applications.

In general, labs are to be completed with a partner, with responses to the embedded questions due on Mondays. You and your partner can expect to receive roughly 45 minutes of class time each day to work through these labs, which should be sufficient to come close to finishing them. This in-class time investment is very important to succeeding in this course, and you should expect check-ins with the instructor to ensure you’re making good progress.

Week 1 (1/19 and 1/21)
- R programming fundamentals
- Submit your answers via Canvas by Monday 1/25 at 11:59pm
Week 2 (1/26 and 1/28)
- Data Exploration in R
- Submit your answers via Canvas by Monday 2/1 at 11:59pm
Week 3 (2/2 and 2/4)
- Simple linear regression
- Submit your answers via Canvas by Monday 2/8 at 11:59pm
Week 4 (2/9 and 2/11)
- K-nearest neighbors and model comparisons
- Submit your answers via Canvas by Monday 2/22 at 11:59pm
Week 5 (2/16 and 2/18)
- Project #1 meetings and presentations (no lab this week)
Week 6 (2/23 and 2/25)
- Introduction to multiple regression
- Submit your answers via Canvas by Monday 3/1 at 11:59pm
Week 7 (3/2 and 3/4)
- Multiple regression - Inference and Variable Selection
- Submit your answers via Canvas by Monday 3/8 at 11:59pm
Week 8 (3/9 and 3/11)
- Multiple regression - Non-linear effects, Interactions, and Influence
- Submit your answers via Canvas by Monday 3/15 at 11:59pm
Week 11 (3/30 and 4/1)
- Logistic Regression - Concepts and Statistical Inference
- Submit your answers via Canvas by Monday 4/5 at 11:59pm
Week 12 (4/6 and 4/8)
- Logistic Regression - Model Assessment and Selection
- Submit your answers via Canvas by Monday 4/12 at 11:59pm
Week 13 (4/13 and 4/15)
- Non-parametric Alternatives to Regression
- Submit your answers via Canvas by Friday 4/23 at 11:59pm

Problem Sets

There will be 4-5 problem sets assigned throughout the semester. These assignments can sometimes be quite lengthy, so if you wait until the last minute to begin them you’ll likely not have enough time.

Problem Set #1
- Due: Friday Feb 12th at 11:59pm
- You should submit your responses via Canvas
- I encourage you to write your solutions in R Markdown (this is not a requirement)
Problem Set #2
- Due: Friday Mar 12th at 11:59pm
- You should submit your responses via Canvas
- Please write your solutions using R Markdown
Problem Set #3
- Due: Friday Apr 16th at 11:59pm
- You should submit your responses via Canvas
- Please write your solutions using R Markdown
Problem Set #4
- Due: Friday Apr 30th at 11:59pm
- You should submit your responses via Canvas
- Please write your solutions using R Markdown

Projects

Project #1 (presented in-class on Thursday 2/18)
- Guidelines - Rubric
Project #2
- Guidelines - Rubric (same as Midterm #1)
Final Project
- Guidelines