Welcome to the TC 310 Home Page

From Mike West's page at Duke University's Institute for Statistics and Decision Sciences:

Assignments

8/25 Read Flip, Introduction and Chapter 1. Start reading Calculated Risks (Finish Part I for Monday). There will be no journal due on Friday, 8/27.

8/27. Do Problem Set #1, which is due on Wednesday, September 1. Remember to do the first part with several other classmates (groups of 3 or 4). However, I want each of you to produce an independent set of coin tosses. Read Flip, Chapters 2-5.

9/3 Do Problem Set #2, which is due on Wednesday, September 8. As usual, this should be done in groups of 3 or 4, and each group should be prepared to discuss its results on Wednesday. Continue reading Calculated Riske, Part 2. Start reading Schmitt's book (the course packet), Chapters 1-3. Try to do the problems Schmitt, and be prepared to ask questions about anything you didn't understand.

9/10 Do Problem Set #3, which is due on Monday, September 13. Again, do this in groups of three or four, and be prepared to discuss your results on Monday. Finish reading Schmitt's book, Chapters 1-3. Bring any questions about the problems to class for discussion.

9/17 Do Problem Set #4, which is due on Monday, September 20. Again, do this in groups of three or rour, and be prepared to discuss your results on Monday. Finish reading Schmitt's book, Chapter 4. Start reading Calculated Risks, Part 3. Be reading Flip, Chapters 18-22.

Here is a study guide for the first test.

10/15 Problem Set #5, due on Monday, October 18. You should have read all of Smart Choices, and this exercise will use some of the ideas in that book. Also, here is the assignment sheet for the group project.

10/22 Problem Set #6, due on Monday, October 25.

10/29 Problem Set #7, due on Monday, November 1.

11/10 Here's a study guide for the second test (10/17)

How to Get Rich Slowly, But Almost $urely!

Modes of Reasoning:

Probability and Inference, Risk and Decision
An Introduction to Bayesian Inference

Life is full of situations where decisions must be made even though the information one has available to make those decisions is incomplete or uncertain, and the consequences of making the wrong decision may be significant. Questions such as: Should I invest in the stock market, and if so, what should I buy, and how much should I invest? If I make the investment, how much can I expect to gain or lose? If I am seriously sick, which of several treatment should I select if they have different side-effects and probability of cure? When sitting on a jury, should I vote that the defendant is guilty or not guilty? As a scientist, should I publish a paper that reports an important new result, even though I cannot be absolutely certain that it is correct? Probability and decision theory can be an important tool in helping us to analyze questions of this sort and make informed decisions. It provides a systematic tool for deciding how our opinions on various issues ought to change as we learn new data. Although the basic principles are very simple, they can be applied in many diverse circumstances, so the tools can be applied to a wide variety of situations. In this course we will investigate how probability and decision theory can help us make important decisions in problems that arise in science, business, the law, medicine, and even daily life.

Readings

The texts for this course are Why Flip a Coin? by H.W. Lewis, Calculated Risks, by Gerd Gigerenzer, Smart Choices, by J.S. Hammond, R.L. Keeney and H. Raiffa, and (relevant portions of) Measuring Uncertainty: An Elementary Introduction to Bayesian Statistics, by Samuel A. Schmitt, which is out of print but is reproduced with permission of the copyright holder and is available in the Texas Union. I will also assign readings relevant to the course taken from newspapers, magazines and journals.

Course Outline

How do we make decisions? We look into the major factors that go into making decisions, both small and large. We use our knowledge to make estimates of numbers, both important and unimportant.
What is probability? We will discuss the concepts of probability first by considering some puzzles and games in which probability plays a role, and then from a more mathematical point of view, discussing the rules of probability. We will also discuss the notion of probability as a way of measuring our degree of belief in a proposition.
Axioms of probability. We will learn the basic rules of probability and how to manipulate the mathematical formulas of probability in simple cases (mostly discrete). We will study the joint and marginal distributions and learn what they mean. We will learn what we mean by coherence and why it is important.
Conditional probability. We will see how probability is conditional on other knowledge that we may have, and is therefore subjective. We will learn the relationship between joint, marginal, and conditional probabilities. We will learn how to use these concepts in practical situations.
Bayes' theorem. We will learn what it is and how to apply it to problems of inference. We will learn about prior probabilities and how to estimate them. We will learn the relationship between prior and posterior probabilities, using Bayes' theorem, and discuss probability theory as a theory of ideal learning. We will discuss the degree to which human beings achieve this ideal. We will discuss the difference between induction and deduction.
Utility and Loss. These tell us how much we stand to gain or lose depending upon which outcome in an uncertain situation is actually realized. We will see how Utility and Loss are related to each other. We will learn how to calculate the expected utility or loss in specific decision problems, and how this informs our decisions.
Decision trees. These are a graphical way of representing the components of a decision problem. We will see how to construct them and how to use them in specific decision problems.
Case studies. We will launch student projects by discussing
* Organizing the project and deciding what question is to be answered
* Where to locate data that would be relevant in making the decision
* How to organize the investigation
* How to present the results
Applications to real-life situations, e.g., investments, jury duty, public policy decision making, gambling.
We will also touch on some other interesting aspects of probability and decision making, e.g., game theory and the Prisoner's Dilemma, paradoxes of probability, the Bayesian Ockham's razor, etc.

Grading

Two one-hour exams, 15% each, no make-ups possible
One two-hour final exam, 15% (Optional, can count as make-up exam)
Problem sets and papers, 20%
Journal 20%
Group Project, 15%
Classroom participation, 15%

When computing the final grade, I will drop the lowest of your three exam grades. This allows you to (a) recover from a bad test or (b) make up a test that you unavoidably missed. It also gives you the option of skipping the final if you don't think that taking it will improve your grade. Because of this policy I will not give make-up exams.

We will have frequent activities and discussions in class illustrating the topics we are studying. There will also be ungraded quizzes and we will discuss the answers to assist you in understanding the material of the course. It is important that you do the problem sets in a timely fashion, since our discussions will often refer to them. I encourage you to work in small groups on the problems, but each student should turn in his or her own write-up. In addition, your small group can act as a study group outside of class.

I will take attendance. Classroom participation is based on attendance as well as participation in the discussions.

Journal

This is a Significant Writing Component course, and a significant part of your grade comes from your writing, including your Journal. You are required to keep a journal. Every week I want you to write in your journal a 3-4 page essay (this means at least three full pages but not more than four pages), to be described below. I want you to use a 12-point proportionally-spaced serif font such as Times Roman, double spaced. This is easier to read than other fonts such as Helvetica or Courier. You should keep your journal in looseleaf form, such as a binder. Please do not use a three-ring type binder...these are bulky and I don't want to have to carry them around. Instead, use a simple heavy paper binder like the one I showed you in class. Insert your weekly essays so that the most recent essay is at the front of the binder. Each week (generally on Friday unless special circumstances obtain) your complete journal to date should be turned in for me to read over the weekend. I will comment on what you have written and return the journal the following week.

Unless otherwise specified, the topic of your weekly journal essay will be to describe in your own words the most interesting or surprising discovery that you made during our discussions in the previous week. If some material was puzzling or even unbelievable, you should address that in your essay as well. Please explain why you chose the particular topics you did, and discuss how the subject of your essay might be important in making decisions. You should write the essay using standard English sentences and paragraphs, with a beginning, a middle, and an ending. I may ask you to revise the essay based on the comments that I make.

I prefer you to turn in your journal in its binder. Under unusual circumstances (e.g., you must unavoidably miss the class when the journal is due) I will accept the essay by E-mail; If you send it by E-mail, please send it as a text file for compatibility reasons. Do not send Microsoft Word or other formatted file formats.

UT Policy Statement

Any student with a documented disability (physical or cognitive) who requires academic accommodations should contact the Services for Students with Disabilities area of the Office of the Dean of Students at 471-6259 (voice) or 471-4641 (TTY for users who are deaf or hard of hearing) as soon as possible to request an official letter outlining authorized accomodations.

Here is a link to my Home Page. From it you can also locate other information about astronomy on the worldwide web. There is information about our department, about McDonald Observatory, and links to other astronomical pages.

This page is under construction. Keep tuned for new material.

This page was served to you by Quasar, a PowerMac. It was last modified on 041110.
My home page is located here. My office number is RLM 16.236. Office hours are MWF 2-3 or by appointment.