Course Description:

An applied introduction to statistical learning and data mining. We learn and demonstrate supervised and unsupervised learning. Some of the topics we cover are clustering, linear regression, model selection and regularization, decision trees and random forests, collaborative filtering, boosting, and methods for evaluation and training. By the end of the quarter, students will:

• Understand the distinction between supervised and unsupervised learning and be able to identify appropriate tools to answer different research questions.
• Become familiar with basic unsupervised procedures including clustering.
• Become familiar with the following regression and classification algorithms: linear regression, ridge regression, the lasso, logistic regression, linear discriminant analysis, K-nearest neighbors, splines, generalized additive models, tree-based methods, and support vector machines.
• Gain a practical appreciation of the bias-variance tradeoff and apply model selection methods based on cross-validation and bootstrapping to a prediction challenge.
• Analyze a real dataset of moderate size using a statistical language such as R or Python.
• Develop the computational skills for data wrangling, collaboration, and reproducible research.
• Be exposed to other topics in machine learning, such as missing data, prediction using time series and relational data, non-linear dimensionality reduction techniques, web-based data visualizations, anomaly detection, and representation learning.

Prerequisites:

• Introductory statistics / probability (preferably at a graduate level, e.g. STATS 116)
• Linear algebra & Multivariable calculus (e.g., Math 51)
• Computer programming (e.g., CS 105)

For your reference, here are some reviews (taken from CS229):

• Review of Probability Theory
• Linear Algebra Review and Reference

Evaluation:

Grades will be based on 3 components:

• 4 homework assignments (worth 40 points each)
• A midterm exam (worth 100 points)
  • Wednesday, July 17, 2019
• A comprehensive in-class final exam (worth 200 points)
  • Saturday, August 17, 2019
  • Optional: A class project (worth 200 points), which can replace the point total of the final exam

If you choose to take the in-class final and submit the class project, we will use the maximum score of the two as your score for the final.

Special arrangements will be made for remote SCPD students who are not able to come to campus for the midterm or the final.

Grades: Grades will be assigned as follows:

98%-100% = A+

93%-98% = A

90%-93% = A-

88%-90% = B+

83%-88% = B

80%-83% = B-

75%-80% = C+

70%-75% = C

60%-70% = D

0%-60% = F

Current point totals will be posted throughout the semester on the course web page.

Homework

You must write up your own solutions individually and explicitly indicate with whom (if anyone) you discussed the homework problems (and what help they provided) at the top of your homework solutions.

Homework problems are similar to writing assignments in other courses in terms of citing sources and plagiarism. Students must cite (via URL or otherwise) sources used in preparing their homework solution.

Programming: While the course textbook uses R, you are free to use whichever (open source) language you like. Your code must be appended to your assignments. You will be graded on the organization, readability, reproducibility, and efficiency of your code. For reference, you can refer to the Google R or Python style guides.

Regrades: Any regrade requests should be submitted through stats202 [at] gmail [dot] com within one week of receiving your grade. Please, read the relevant solutions and review the relevant course material prior to sending a request and specify (1) the part(s) of the homework you believe were wrongly graded and (2) why you deserve additional credit. We will typically regrade the entirety of any homework for which any regrade is requested and the resulting score may be higher or lower than the original one.

Late Assignments: Late homework assignments will only be accepted up to 1-day and incur a 10 point penalty.

Exams:

Technology: A basic hand held calculator which has logarithm functionality will be sufficient for the midterm and the final. To complete homework assignments you will need internet access, Microsoft Excel, and the R statistical software package (free download).

SCPD and other remote students: You're welcome to attend class in person. Regarding the midterm and final exam, please make sure your exam monitor is able to scan and return the exam by e-mailing it back to stats202 [at] gmail [dot] com. I encourage SCPD students to take the midterm and final in class if possible.

Make-Up Exams: I must receive prior notification and justification of your impending absence in order to authorize a make-up exam. Messages must be left either on my cell phone voice mail or sent by email prior to the start of the exam. An exam must be made up within one week of the original exam date. There will be no exceptions.

Academic Honesty: It is essential that you abide by the academic honesty policies of the university. In particular, you may not copy other students' work on exams or homework. Please also keep in mind the university honor code.