Statistical Inference is a branch of Mathematical Engineering that studies ways of extracting reliable information from limited data for learning, prediction, and decision making in the presence of uncertainty. The main goals of this course are:
• Develop statistical thinking and intuitive feel for the subject,
• Introduce the most fundamental ideas, concepts, and methods of Statistical Inference, and
• Explain how and why they work, and when they don’t.
If you do well in the class, you should be able to read (and understand) most contemporary papers that use statistical inference and perform statistical analysis yourself.

This is an introductory course on statistical inference. No prior knowledge of statistics is assumed. However, a solid understanding of Probability is required. Ma 3 or ACM/EE/IDS 116 (or equivalent) is a “hard” prerequisite. A key part of the course is problem sets, where you will get experience in using the learned methods and models in applications via simulations in MATLAB. So, some familiarity with MATLAB (and programming in general) is desired, but this is a “soft” prerequisite: MATLAB is easy to pick up on the fly, especially for the purposes of this course.

There is not a single book the course is based on (I am writing the one!). Good news: I will provide comprehensive lecture notes. After each lecture, I will be uploading the corresponding notes to the course Piazza page together with supplementary materials for further reading. The course was developed using the following books, which can be used as supplementary (but not required) textbooks:
• G. Casella & R.L. Berger, Statistical Inference, 2002.
• A.C. Davison, Statistical Models, 2003.
• L.A. Wasserman, A Concise Course in Statistical Inference, 2005.
• M. Lavine, Introduction to Statistical Thought, 2013.
• S.L. Lohr, Sampling: Design and Analysis, 2010.
• D.C. Montgomery, E.A. Peck, & G.G. Vining, Introduction to Linear Regression Analysis, 2006.
• D. Nolan & T. Speed, Stat Labs: Mathematical Statistics Through Applications, 2000.
• S. Weisberg, Applied Linear Regression, 2005.

Your final grade will be based on your total score. Your total score is a weighted average of Problem Sets (60%), Midterm exam (20%), and Final exam (20%). You can increase your total score by up to 5% if you participate actively in Piazza discussions in the Q&A section. Every answer submitted before TAs or instructor answer, which is later endorsed as “good answer” by TAs or instructor, gets 1% of the total score. There are no fixed thresholds for grades, but if your total score is 90% (80%, 70%, 60%), you are guaranteed at least “A” (“B”, “C”, “D”).

There will be six Problem Sets. Problems (and solutions) will be posted on Piazza. For assignment and due dates see “Important Dates” below. Late submissions will not be accepted for any reason, but the Problem Set with the lowest score will be dropped and not counted toward your total score. Submitting wrong files or files in a wrong format is considered as a late submission. Extensions may be granted for academic, personal, or medical reasons. For extensions, please email the Head TA.