- Instructor:
- Bruce Draper
Office: 442 CS Building
Office Hours: By arrangement
Email: draper@colostate.edu
- GTA:
- Dejan Markovikj
Office:
Office Hours:
Email: cs540@cs.colostate.edu
- Lecture Time and Place:
- 11:00 - 12:15, Tue, Thurs, CSB Room 425
Artificially intelligent agents perceive the world, solve problems by combining data, knowledge, and judgement, and act in the world. Examples of AI agents range from web search engines to scheduling agents to humanoid robots. This course will cover representations and algorithms in several core subareas of artificial intelligence, beginning with advanced search and planning. We will then tackle more advanced topics selected in response to student interests. Candidate topics include (but are not limited to): evolutionary computation and particle filters, ensemble learning, Bayesian networks, Markov and other time-series models, data mining, information retrieval, and natural language interpretation/generation.
The class is structured around a lecture format, individual programming assignments, team projects, and a midterm. Class
discussions, questions and participation (in class or via online discussions) are strongly encouraged.
Pre-requisites:
CS440 or equivalent. Knowledge of fundamentals of Artificial Intelligence, search and logic. Roughly the material covered in the first ten chapters of Russell & Norvig.
Grading:
Students will be graded based on individual programming assignments, a midterm, and team programming assignments. The individual programming assignments, of which there will be either three or four (depending on how things go), are traditional programming assignments in the sense that a specific task is assigned, as the students are expected to write a program that satisfies the task. Students will be graded on how well their program performs, assuming they have met all other expectations (for example, that all the submitted work is their own). The midterm will be a one hour, fifteen minute exam given in class. The team programming assignments (probably two) are different. The project topics will assigned in very broad terms. The teams are expected to do interesting work within the assigned parameters, and then write a paper and give a short oral presentation in which they justify their specific approach, compare it to the material presented in class, and tell us what they have learned. Every project particpant must also write a paragraph outlining their contributions to the project, and not every member of a team will necessarily receive the same grade for a project. Grades will be determined as follows:
| Activity |
Weight |
| Programming Assignments | 30% |
| Midterm | 30% |
| Team Projects | 40% |
Materials
To train students to read the literature, reading assignments will be journal and conference papers, the links to which will be posted online. The written portion of team projects are expected to including the reading assignments in their literature reviews, and compare and contrast the paper with their projects.
Late Policy:
There is no excuse for being late, and late projects will not be accepted.
If something radical happens in your life that makes it impossible to make the deadline (e.g., a family member is rushed to intensive care and you need to be there or better yet you win mega-millions and have to be holed up with lawyers for the next 4 days), contact the instructor ASAP. However, the assignments are expected to take some time, so the instructor will have little sympathy if you had not yet started it and you are within 48 hours of the deadline.
Class Participation
All students taking this course are expected to participate actively. This includes asking and responding to questions. For distance students, the mechanism for asking and responding to questions is Piazza.
Students registered in the on-line section will be able to watch video recordings of lectures. These will be available through Canvas.
Office Hours:
Generally, I have found that graduate students prefer arranging times at our mutual convenience. However, for the DL students, I can set a time each week that I will be available through the chat facility on RamCT. In any case, if you want to talk outside that time, send an email to arrange a time.
