Going From C.S. Undergrad to E.E. Graduate

[rgreen59]

Hello,

Sorry if this question has been asked before. I have spent several hours googling for an answer, but have yet to find an answer. I came across similar questions at physics forums and thought this would be a good place to ask.

Basically I began my career in EE (in the military prior to going to college) and I wanted to write software for analog and digital communications software. The team I worked with consisted of EE, CE, Math, and CS majors. I decided to leave the military and attend college. Because I still had to work, I attended night courses in Computer and Information Science (CMIS). Upon graduating, I quickly learned that nobody in the EE industry really wanted me because their EE majors could program and they new all of the EE stuff as well.

I want to go back for a M.S. degree in EE. I am currently going back and taking Calculus and Linear Algebra courses at the community college because I know these are required to understand EE. I am willing to take other necessary courses as well, but I am on a limited time schedule. Is my desire to get an M.S. in E.E. with a C.S. undergraduate degree even possible? Has anyone else with an undergraduate degree in C.S. or C.M.I.S. went back for a masters in EE? All advice is greatly appreciated.

Although, I believe a EE degree would hold more weight, there is an Applied Math program at a local university that allows half of the courses to be taken in a math centric program like EE. While this is the best option I have found so far, I would really prefer to go for a pure EE degree.

Thank you in advance,
Ryan

 

[Maxwell]

I'm working on my MS in EE right now, so I'll give you my opinion from that angle.

I think there is a pretty big difference between a CS and a CMIS degree. The CS degree usually involves more math - at least up to Calculus 2. Since you are taking these courses already (or you will, in the case of Calc 2), you seem to be on the right track.

If you want to pursue an MSEE, you're going to need a class on differential equations as well. Since you were writing software for communication programs, I'm assuming you have had at least basic classes in probability and signals?

I have a friend who did his BS in Physics and is now in grad school for EE. The department made him take a bunch of intro classes like electronics (he got to skip basic circuits, though), signals & systems, and a few design classes. Depending on what field of EE you go into you will have to take the intro classes in that field - like control systems, communication systems, intro to VLSI circuits, etc.

It's certainly possible to get your MS in EE if you've done CS, but certain fields might be easier for you to enter than others. Since you've mentioned you are on a limited time schedule, you might want to consider entering a field of EE in which you are already familiar with the basic concepts. Have you thought about what field in EE you'd like to go into?

 

[rgreen59]

Maxwell,

Thank you for the quick reply. I do agree there is a difference between CS and CMIS. CMIS was my only option for night classes, but I did try to take the hardest classes within the program. The CMIS program only required discrete math and by the time I realized I should take calculus, I already had too many lower level courses. That is why I am going back to take these courses now. I did well in calculus I, but haven't taken calculus II yet. Instead I am taking linear algebra this term. Thank you for your advice on differential equations. How about Calculus III? Should I take it or could I get by without it?

My training in signals was taught in a six month military course titled Communications and Signals Processing as well as through 7 years of on the job training. I do not know how this training translates to academia. I was mostly taught different modulation, keying, and multiplexing techniques. Everything I learned was communication signals related.

The two areas of EE I seem primarily interested in are communications and signal processing. I feel I understand the communications stuff pretty well, but my understanding of signals processing (in a mathematical sense) is weak. I recently picked up a book on signals and systems and I was quickly lost, but I am still determined to continue trying to learn.

Off the topic, but can you please tell me the difference between signals and systems and digital signal processing?

Thanks again,
Ryan

 

[Maxwell]

Hey Ryan,

I understand your situation with the CMIS course timings. Sometimes you take whatever you can. Regarding Calculus III - I believe that every ABET accredited undergraduate EE degree program requires it. I took it my sophomore year of undergrad. You're best bet is to take it while you're taking classes at the community college. You'll get what you need for engineering out of it and it won't be so heavy on the proofs. And again, definitely take a differential equations class.

You seem to have done a lot of basic communications theory. An intro comm. course will do a bunch of different modulation techniques, filters, etc. However, if you get into either DSP or Comm it will get extremely heavy on the math. Both fields rely big time on probability theory and linear algebra.

Signals & Systems is usually just the name for the first signal processing class that EE undergrads take. It just deals with continuous and sometimes discrete signals and how those signals are processed through various systems. This class typically covers two major tools that EE's use all the time - Fourier Transforms/Series (and the discrete Fourier Transform) and Laplace Transforms (and Z-Transforms in discrete). Have you ever heard of these? They are two of the major cornerstones upon which DSP and Comm theory (among many other EE fields) are built upon.

DSP is a big field of EE that takes the ideas from that basic signals and systems class and builds upon them. Different DSP classes might be on compression theory, estimation theory, adaptive signal processing, etc. In my experience, I've noticed a great amount of overlap between DSP and Comm. courses. They do break of into a bunch of different subfields, but you will take plenty of courses in each field if you decide to specialize in one of them.

A lot of people who go into DSP or Comm. like staying on the theoretical side of EE. I've noticed a lot of my friends who are focusing in DSP or Comm. don't care a single bit about how their ideas and algorithms are actually going to be (or can be) implemented - they just like creating/learning new techniques to garner information from signals, send signals, or filter signals.

Regarding the intro signals & systems textbook - you should be able to barely understand the material if you've taken Calculus I. However, it's really better if you've taken some more math first. Did you begin to study things like convolution? That's usually covered in the beginning of a S&S textbook. You wouldn't really start to study Fourier Transforms until they've covered the introductory frequency domain concepts in the book.

 

[rgreen59]

Maxwell,

You’ve provided great information very quickly and I appreciate it. I know that I am off course and that is why I am seeking advice.

If you were in my position, what courses would you take before applying to an EE graduate program with an undergraduate CMIS degree? Based on my research and your advice, I will take calculus I, II, and III, linear algebra, and differential equations. I was also thinking I should take physics I and II and maybe statistics. Do you recommend all of these as well? Again, I want to start graduate schools as soon as possible so I would like to take as few courses as possible without jeopardizing my graduate school performance.

Regarding you questions in your reply, yes I have heard of the different transforms and have read about them on Wikipedia. I believe I understand conceptually the difference between time and frequency domains and my understanding is that these transforms convert signals from one domain to another. Is that right? Also, I am not familiar with convolution as of yet. I did a quick search on Wikipedia and it didn’t appear too foreign, but it appears to be heavily based on integrals, which I haven’t studied in-depth yet.

Thank you again,
Ryan

 

[Maxwell]

Hey Ryan,

If I were in your position I would try to get as many non-EE courses done at community college as I can. I would definitely get in touch with the university I plan to apply to for graduate school (both the main grad school dept. AND the EE dept.) and set up a meeting. I would go over the prerequisites that I should have under my belt before applying, and I would get the non-EE courses (and perhaps the very basic EE courses) done at my community college. If you have not taken physics yet, you will certainly need that under your belt before you apply to EE grad school.

You have the general idea of the transforms down. EE's use the FT to represent a time-domain signal in the frequency-domain. In the frequency domain we have a lot of useful techniques and properties we can exploit. If you have not covered the basics of signals & systems (like convolution), then you do not need to worry about this yet.

Do make an appointment to talk to a transfer councilor at the university you plan to apply to, though. That is something you should do so you can start planning the next few semesters.