UMass CICS - COMPSCI 383

COMPSCI 383 (Fall 2020)

Greetings and Salutations

Welcome to 383! We are excited to have you enrolled for the coming semester, and we hope to create a learning environment where each individual feels recognized, supported, and celebrated. Please know that regardless of your background, age, race, ethnicity, geographical origin, citizenship status, language, sex, gender or gender identity, sexual orientation, disability, family status, religion, political views, military experience, socioeconomic status, or education, you are welcome in this class! Furthermore, the ideas and experiences that each one of you brings to our (virtual) classroom are crucial to our success. Thank you for being here.

Below you'll find details on the course organization, along with policies on conduct, grading, and communication. Please read it over in its entirety, and let us know if anything is unclear.

[ Last updated: 22 August 2020 ]

Course Description

Objectives
This course focuses on understanding the core concepts and algorithms in AI. Over the course of the semester, students will gain an understanding of the foundational theoretical concepts of aritificial intelligence, including search, knowledge representation, machine learning, and formal logic. We will learn how these concepts are applied in the context of several applications, and discuss the societal impacts of different approaches wherever possible.

Learning Outcomes
After completing the course, students should be able to:

Formulate tasks as search problems, including state space descriptions, transition functions, and goals.
Understand the role of heuristics and evaluation functions in search problems, and use them to find solutions.
Identify different types of machine learning problems (e.g., supervised vs. unsupervised) and be able to suggest appropriate approaches and algorithms for solving them.
Develop AI solutions using the Python programming language.

Prerequisites
Formally, the prerequisites for 383 are COMPSCI 220 or COMPSCI 230, and COMPSCI 240 or STAT 515. Informally, this means you are expected to come in knowing how to structure and develop computer programs, and have an understanding of the basic concepts of probability theory. Note that assignments for the course will be written in Python, and while we will provide resources for getting up to speed, learning the language will not be covered.

Schedule
The course schedule can be found here.

Course Policies

General Code of Conduct
The course Code of Conduct can be found here. Learn it, know it, live it.

Homework Assignments and Grading
Your final grade will be calculated from your scores on homework assignments and exams, using whatever percentages within the following ranges will maximize your grade:

50-80% Homework Assignments
10-25% Midterm exam
10-25% Final Exam
+0-3% Participation (based on helpful use of class forums, office hours, and other forms of engagement)

Homework Lateness Policy

All assignments should be submitted through Gradescope. You may submit one assignment up to 3 days late without penalty. After that, any assignments submitted late will be subject to a 15% penalty per day (e.g., an assignment submitted 26 hours late will have its final score reduced by 30%). No assignments will be accepted more than 3 days late. You may request extensions due to serious and documented medical or family emergencies.

Homework Bug Bounties

As unimagineable as it may seem, homework assignments may occasionally contain errors or bugs in the supplied code. The first person to identify an error or bug in the code and post about it to the class forum will receive a +5% bonus for their grade on that assigment. The first person to post an elegant fix to the problem will receive +1-10% bonus (depending on the difficulty of crafting the solution or workaround).

Academic Honesty and Collaboration Policy

You are encouraged to discuss the programming projects with your peers, and may optionally partner with one other person to code/write up your solutions. Both students are expected to be equal contributors, and are responsible for writing, understanding, and/or testing your solution. For coding projects, I highly recommend pair programming.

In addition, we follow the University's Academic Honesty Policy and Procedures. In particular, the following are prohibited:

Misrepresenting code written by others as your own
Sharing or making available your solutions to anyone other than your partner
Posting or selling class materials or solutions online
Claiming someone as a partner when they were not an equal contributor to the work

All of the above is subject to change.