<p>I'm been trying to find colleges with a sizable number of courses in AI for undergraduates (at least 10), but have yet to find any. When looking at Berkeley's courses, I only found 3 on AI.</p>
<p>What colleges have a sizable number of courses in AI? Why doesn't Berkeley have more? Do any colleges have a sizable number of courses in AI? If not, there where do AI users and researches get educated? Are most AI courses just reserved for graduate students?</p>
<p>AI is a subarea within CS, so there may be just one or a few upper division undergraduate courses in it at most schools. Those CS departments with a strong emphasis on AI would have a greater number of graduate level courses and most research opportunities in AI.</p>
<p>Thanks for the response, but that doesn’t really answer my questions. I thought Berkeley did have a strong emphasis in AI, as it was listed in “universities with AI programs” in AI international and was tied as the top schools in AI by US News. I know US News’s rankings are flawed, but I’d think Berkeley would have more than just 3 courses. Why doesn’t it have more courses? Am I just looking in the wrong places? Do people looking to learn AI in more specific fields just help others do research in it?</p>
<p>MIT has a lot of AI courses - something like 8 or 9 undergraduate and more than 10 graduate. UMass has a fair amount : 6 or 7. Check out their undergraduate AI track. I’m sure that Stanford and CMU have a lot, and there must be others that have a decent amount.</p>
<p>However, even at these schools you’d find it hard to fit more than about 3 AI courses into your CS program. A typical undergraduate focusing on AI at one of these schools would probably take 2 - 4 AI courses (depending on how much flexibility they have) and do their undergraduate research and thesis on an AI topic.</p>
<p>Why would it be hard to fit more than 3 AI courses in one’s CS program? Don’t most students take much more courses than that? Also do students learn more in-depth information about AI when doing their undergraduate research and thesis?</p>
<p>Well, I’m assuming you’ve seen the list of courses on this page:
<a href=“http://www.eecs.berkeley.edu/Research/Areas/AI/”>http://www.eecs.berkeley.edu/Research/Areas/AI/</a></p>
<p>CS 188. Introduction to Artificial Intelligence
CS 280. Computer Vision
CS C281/C281A. Statistical Learning Theory
CS 281B. Advanced Topics in Learning and Decision Making
CS 287. Advanced Robotics
CS 288. Artificial Intelligence Approach to Natural Language Processing
CS 294. Artificial Intelligence Programming Techniques
EE C125/C215A. Introduction to Robotics
EE 128. Feedback Control Theory
EE 227A. Introduction to Convex Optimization
EE 290P. Advanced Topics in Bioelectronics
EE/CS 298. Center for Intelligent Systems (CIS)</p>
<p>And here’s the general catalog of courses:
<a href=“http://general-catalog.berkeley.edu/catalog/gcc_list_crse_req?p_dept_name=Electrical+Engineering+and+Computer+Sciences&p_dept_cd=EECS&p_path=*”>http://general-catalog.berkeley.edu/catalog/gcc_list_crse_req?p_dept_name=Electrical+Engineering+and+Computer+Sciences&p_dept_cd=EECS&p_path=*</a></p>
<p>You can take the 200 level stuff, but it looks like you need instructor approval.
<a href=“Getting Into Computer Science Classes - Google Docs”>Getting Into Computer Science Classes - Google Docs;
<p>Here’s Stanford’s stuff: <a href=“Courses | Stanford Artificial Intelligence Laboratory”>http://ai.stanford.edu/courses/</a>
…which doesn’t really seem to be much different. They also list three undergraduate classes (one is really a philosophy course). Some of the courses Berkeley has that not listed for Stanford and vice versa are probably offered, just listed elsewhere.</p>
<p>U Mass has a special track within their CS degree (<a href=“https://www.cs.umass.edu/ugrad-education/artificial-intelligence”>https://www.cs.umass.edu/ugrad-education/artificial-intelligence</a>), which list the following courses (I’m not sure what the numbering means):</p>
<p>Three required courses:
311 Algorithms
320 Software Engineering
383 Artificial Intelligence </p>
<p>Any two courses from the following:
370 Computer Vision (or 670)
403 Robotics<br>
446 Search Engines*<br>
503 Emb. Systems
585 Natural Lang. Proc.<br>
589 Machine Learning<br>
683 Advanced AI </p>
<p>And really, I doubt any of these courses are much different than what you’ll find at Berkeley or Stanford.</p>
<p>Northwestern and Yale CS are known for their AI focus.</p>
<p>@XDSFO904 - I suggest you answer your own question by picking a couple of the above schools and working out a 4 year course schedule for each one, making sure you cover university requirements, department requirements, and prerequisites. See how many AI classes you can fit in. The CS web pages at MIT, for example, have detailed planning information, and I assume most of the other schools do as well. </p>
<p>Most undergrad CS programs are concerned with giving you a broad, general introduction to CS. I find it hard to imagine an undergraduate CS program where you’d be able take 10 AI classes. I’d think 3 or 4 classes would be about as good as it gets.</p>
<p>If you really want to get into AI, you’ll likely have to go to grad school.</p>
<p>You might want to post the question in the CS and Math forum.</p>
<p><a href=“Math/Computer Science Majors - College Confidential Forums”>http://talk.collegeconfidential.com/math-computer-science-majors/</a></p>
<p>@xdsfo904:</p>
<p>Tufts is big in AI and has an interdisciplinary/application/undergrad focus, so there are 13 AI courses </p>
<p>4 are accessible after taking 3- 4 math courses and 2 CS courses
after taking 3 more CS courses, 6 more are available
after taking one more CS course 1 more becomes available, or after taking two more math courses 1 more becomes available.
1 is available by special permission only.</p>
<p>Then there is also your senior design project (which is equivalent to 2 courses) that you could do on an AI related topic.</p>
<p>The courses are not offered every semester, but with careful planning, you should be able to get in more than 4, especially if you have AP credit.</p>
<p>You can also take some of your liberal arts courses in Cognitive/Brain Science related areas.</p>
<p>Computer Science can be taken through the College of Arts and Sciences or the College of Engineering. If you wanted to minimize the number of other requirements, you can design your own major in the College of Engineering, but it will not be ABET approved (which is OK if you know what you are doing and get input from experienced people). </p>
<p><a href=“Prospective Students | Department of Computer Science”>Prospective Students | Department of Computer Science;
<p><a href=“Untitled”>Untitled;
<p>There is also CS related research you could participate in. The just landed funding for a multi-university project to attempt to teach morals to robots:</p>
<p><a href=“http://hrilab.tufts.edu/research/”>http://hrilab.tufts.edu/research/</a></p>
<p>Cambridge/MIT (a few subways stops away) also has an active AI community and lots of startups.</p>
<p><a href=“New England Artificial Intelligence (Cambridge, MA) | Meetup”>http://www.■■■■■■■■■■/intelligence/</a></p>
<p>The strongest programs are most likely in the Valley or Boston (the two largest computing clusters).</p>
<p>MIT/Harvard/Tufts are right on the Red Line subway (which is becoming the new high-tech/start-up belt) so there are lots of opportunities for internships (even during the school year).</p>
<p>
</p>
<p>“AI” is a term that was used a lot back in the 1980s (and before). Now that 2001 has come and gone, we still don’t have HAL (although Siri is pretty good). Technologists may be more realistic about what can be achieved, and at the same time more focused on specific functions that mimic intelligence in a useful way (such as speech processing or robotics).</p>
<p>Look for course offerings in areas such as Natural Language Processing, robotics, computer vision, expert systems, machine learning, etc. </p>
<p>Thanks bigwords, I never actually found that list. How did you find it?</p>
<p>Thanks for all the responses. Also, how important is participating in AI research as an undergraduate as opposed to just taking classes in it?</p>
<p>AI is better done as a graduate degree. That said, you can be CS major and do research with a professor in AI. How important is it? VERY, VERY, important. Would you trust someone who says they can drive a car because they’ve read alot of books about it? I’m guessing you’d rather have some one who’s actually tried driving one first.</p>
<p>Thanks, IrishIs. Should one partake in AI research as an undergraduate or a graduate?</p>
<p>tk21769, I did try to find courses about subjects about AI that aren’t termed “artificial intelligence.”</p>
<p>I find the Berkeley situation puzzling. In general, one could argue that any CS specialization is “better done” at the graduate level (because there are more courses), but it shouldn’t be required. </p>
<p>UMass was a pioneer in AI. Back in the early 80’s the Intro level AI course was “500 level” and AI was considered an area for graduate-level study. Now the AI course is “300 level” and it is considered an area for an undergrad specialization (hence the track). There are more AI courses offered at UMass than listed above, just check the course list. 100 and 200 level courses are introductory level. 300 and 400 level courses are intermediate level. 500 and 600 level courses are advanced/graduate level, but undergrads can take many of them (there tend to be more prerequisites, though). </p>
<p>Within the last decade, AI techniques have become more mainstream and they are being applied to a number of disciplines. I think it makes a for a great area of specialization at the undergrad level if you are looking to apply the techniques in novel ways. If you are looking to advance the state of the art within AI, that is best done at the graduate level. If you want to apply to a Phd program, then undergrad research is very important. If you want to go into industry, then special projects and internships are important, because a recruiter will want to see that you can actually apply what you learned in a useful way.</p>
<p>Tufts started a major expansion of Comp Sci about a decade ago (which is too recent to show up in any US News Survey) and decided to focus on interdisciplinary AI research rather than research within the AI discipline. This resulted in a high percentage of professors with AI expertise and a lot of AI related courses (relative to the size of the department). Another perusal of the course catalog revealed a few more courses than referenced above. Two more comp sci professors with machine learning expertise and a cog sci/machine learning professor are expected to be hired within the next year.</p>
<p>Some schools chose to focus on research within the discipline rather than interdisciplinary research, and some schools tend to focus on their research rather than developing new courses based on their research, both of which would reduce the variety of courses available. This doesn’t make the program any better or worse, just different. You need to find a program that best meets your interests and objectives.</p>