<p>then i’m gonna add 3 physics upper div. courses: 110, 137, 141</p>
<p>Correct me if I’m wrong, but some of those EE/CS classes are not going to be related to AI in any way. For example, EE122 is an Internet networking protocol class (Ethernet, TCP/IP, etc…) which is not going to be useful at all for AI (the “networking” designation may be a little misleading).</p>
<p>CS 186 is also a graphics class. From the syllabus, you’ll be doing graphics rendering stuff, which is not very much related to AI or even computer vision. Someone who would take this class would be, for example, someone who wants to work at Pixar or a video game company.</p>
<p>EE 123 also does not sound <em>that</em> useful for AI, looking at the syllabus. (Although it may be a <em>tiny bit</em> useful, which is not enough to make it worthwhile, IMO).</p>
<p>And some of these classes are not going to be required to grad school, either, IMO.</p>
<p>EE 122 (networks) and CS 186 (databases) will contain concepts that are commonly used. Computers frequently need to communicate data with each other (networks) and store/organize it efficiently (databases), regardless of the application.</p>
<p>Actually, CS 280: “This course is an introductory graduate course in computer vision. We will cover principles of image formation, local feature analysis, multi-view geometry, image warping and stitching, structure from motion, and visual recognition. We will also touch upon related topics in signal and image processing including convolution and image pyramids, and may cover computer graphics topics involving computational photography and image-based rendering as time permits.”</p>
<p>So graphics and digital signal processing will both have some kind of help in Computer Vision if I ever decide to go to grad school in it.</p>
<p>Technically the more different mathematical “views” you have of any topic in life, I’d say the more “unique” and “diverse” solutions you can come up with to solving a problem than has yet to be solved or to solve an existing problem in a much more efficient manner. </p>
<p>For now I seem to be interested in AI, but I believe understanding various algorithmic techniques (whether in Digital Signal Processing or graphics or Databases or in the CS theory classes), how communication works (internet class), or robotics will give me different mathematical “views” or “perspectives” when deciding to come up with a solution to a problem. [This is why I’d really like to take EE126 too to understand continuous probability and random processes better for the machine learning side of AI, but I have no space]</p>
<p>I could be completely wrong though.</p>
<p>^Actually I think CS184 actually isn’t all that useful for what I want, it just seemed cool to know the theory behind 3D computer graphics. It seems EECS 149 (Embedded Systems) and EE 128 (Feedback Control) are more useful courses than CS184 or CS164 in terms of robotics and CS systems that interact with the physical world (autonomous machines, etc)</p>
<p>Sigh I give up. I think I finally accepted that learning everything is impossible I should just be more realistic in what I want…</p>
<p>Looking at the graphics course syllabus, it’s all about 3d rendering, ray tracing, etc. I dont think there is much overlap (my opinion).</p>
<p>E122, networking, focuses on IP, DNS, QoS, Ethernet, TCP, all of which are internet infrastructure protocols, none of which are topics that IMO are related to robotics or AI. These are topics you would study if you work with servers, routers, switches, etc…
That is not to say this class is not useful. It’s just not necessary.</p>
<p>I also think you need to consider whether you’re leaning more towards AI (software-oriented) or robotics (more hardware/electronics-oriented), and choose accordingly to emphasize CS or EE.</p>
<p>Again, rather than just going off the course title, read the syllabus, look at the lectures, ask people who’ve taken the class. For all you know, the type of feedback control the class refers to is not the one you have in mind.</p>