Course Description:  A study of statistical and theoretical foundations of modern statistical machine learning. Topics covered include: resampling methods (cross-validation and bootstrap), model evaluation and variable selection, maximal margin classifiers and support vector machines, tree-based methods for regression and classification, ensemble methods, mixture models, and unsupervised learning,

Prerequisite(s):  Advanced knowledge of R programming.  Graduate level mathematical statistics courses at the level of 6201 and 6202. Graduate level linear model course at the level of 6214.

Required Text:

The Elements of Statistical Learning, 2nd  Edition by Hastie, Tibshirani & Friedman.

Available online: www.web.stanford.edu/~hastie/ElemStatLearn.

Recommended Texts:

Pattern Recognition and Machine Learning  (2006) by Bishop.*

An Introduction to Statistical Learning, 2nd  Edition by James, Witten, Hastie & Tibshirani.

Available online: https://hastie.su.domains/ISLR2/ISLRv2_website.pdf.

Videos and notes: https://tdg5.github.io/stats-learning-notes/.

Learning Outcomes: At the completion of this course students will be able to: (1) formulate the statisti- cal/probabilistic structure of common machine learning techniques, (2) understand theoretical aspects of the design and evaluation of statistical learning methods/algorithms, (3) identify appropriate statistical learning methods for different types of data, and (4) program, evaluate, and interpret statistical learning methods and algorithms in R.

Grading:

Homework Assignments    65%

Homework Presentation    5%

Special Topic Lecture        30%

● Students are expected to attend all lectures during class time (2.5 hours a week) either in-person or virtually.

● Students are expected to spend a minimum of 10 hours a week on out-of-class/independent learning.

Mode of Instruction:

Lectures will begin as virtual class meetings on Blackboard Collaborate Ultra (BBCU). Lectures will be in- person starting the second week of class, but will be synchronously streamed online through BBCU for students who have university approval to attend virtually.  You are permitted to attend a given lecture virtually on a case-by-case basis with approval from me in advance. You are expected to attend all lectures in-person or virtually (if approved), but lectures will be recorded and posted in case virtual attendees experience technical difficulties and/or internet connectivity issues.

Blackboard:

All students are required to register for the course in Blackboard, the GWU web-based instructional resource. Course information and materials, including homework assignments and grades will be posted there. It is the students responsibility to check the Blackboard website frequently for up-to-date information.

Use of Electronic Course Materials and Class Recordings:

Students are encouraged to use electronic course materials,  including recorded class sessions,  for private personal use in connection with their academic program of study.  Electronic course materials and recorded class sessions should not be shared or used for non-course related purposes unless express permission has been granted by the instructor.  Students who impermissibly share any electronic course materials are subject to discipline under the Student Code of Conduct. Please contact the instructor if you have questions regarding what constitutes permissible or impermissible use of electronic course materials and/or recorded class sessions. Please contact Disability Support Services at www.disabilitysupport.gwu.edu if you have questions or need assistance in accessing electronic course materials.

Computing:

We will use R for statistical computing. Please download the R software online at: www.r-project.org and www.rstudio.com.

Academic Integrity Code:

Academic Integrity is an integral part of the educational process, and GW takes these matters very seriously. Violations of academic integrity occur when students fail to cite research sources properly, engage in unau- thorized collaboration, falsify data, and in other ways outlined in the Code of Academic Integrity.  Students accused of academic integrity violations should contact the Office of Academic Integrity to learn more about their rights and options in the process. Outcomes can range from failure of assignment to expulsion from the University, including a transcript notation.  The Office of Academic Integrity maintains a permanent record of the violation.

More information is available from the Office of Academic Integrity at www.studentconduct.gwu.edu/code  - academic-integrity.  The Universitys “Guide of Academic Integrity in Online Learning Environments” is available at www.studentconduct.gwu.edu/guide-academic-integrity-online-learning-environments. Contact information: [email protected] or 202-994-6757.

Safety and Security:

● In an emergency: call GWPD 202-994-6111 or 911.

● For situation-specific actions:  review the Emergency Response Handbook at www.safety.gwu.edu/ emergency-response-handbook.

● In an active violence situation: Get Out, Hide Out, or Take Out. See www.go.gwu.edu/shooterprep.

● Stay informed: www.safety.gwu.edu/stay-informed.

University Policy on Observance of Religious Holidays:

In accordance with University policy, students should notify faculty during the first week of the semester of their intention to be absent from class on their day(s) of religious observance.  For details and policy, see “Religious Holidays” at www.provost.gwu.edu/policies-procedures-and-guidelines.

Support for Students Outside the Classroom:

Virtual Academic Support

A full range of academic support is offered virtually in Fall 2020. See https://coronavirus.gwu.edu/top  - faqs for updates. Tutoring and course review sessions are offered through Academic Commons in an online format. See https://academiccommons.gwu.edu/tutoring. Writing and research consultations are available online.  See https://academiccommons.gwu.edu/writing-research-help.  Coaching, offered through the Office of Student Success, is available in a virtual format. See https://studentsuccess.gwu.edu/academic  - program-support. Academic Commons offers several short videos addressing different virtual learning strate- gies for the unique circumstances of the Fall 2020 semester. See https://academiccommons.gwu.edu/study  - skills. They also offer a variety of live virtual workshops to equip students with the tools they need to suc- ceed in a virtual environment. See www.tinyurl.com/gw-virtual-learning.

Writing  Center

GWs Writing Center cultivates confident writers in the University community by facilitating collaborative, critical, and inclusive conversations at all stages of the writing process. Working alongside peer mentors, writ- ers develop strategies to write independently in academic and public settings.  Appointments can be booked online. See https://gwu.mywconline.

Academic  Commons

Academic Commons provides tutoring and other academic support resources to students in many courses. Students can schedule virtual one-on-one appointments or attend virtual drop-in sessions.   Students may schedule an appointment, review the tutoring schedule, access other academic support resources, or obtain assistance at https://academiccommons.gwu.edu.

Disability Support Services  (DSS): 202-994-8250

Any student who may need an accommodation based on the potential impact of a disability should contact the Disability Support Services to establish eligibility and to coordinate reasonable accommodations.  For additional information see: www.disabilitysupport.gwu.edu.

Counseling and Psychological Services :  202-994-5300

GWs Colonial Health Center offers counseling and psychological services, supporting mental health and per- sonal development by collaborating directly with students to overcome challenges and difficulties that may interfere with academic, emotional, and personal success. For additional information see: www.healthcenter. gwu.edu/counseling-and-psychological-services.