Course Choice: Robotics, ML, Adaptive Signal Processing

[Qurks]

I wasn't sure if I should post this here or academic guidance but given the career factor in this question I will post it here:

I am doing a masters in electrical engineering with a focus on signal processing but I am also interested in others areas such as machine learning and robotics. I need to decide what courses to take this coming year.

I will need to pick two out of the following three:

1. Adaptive Signal Processing
2. Machine Learning.
3. Intro to Robotics.

I'm wondering, will the intro-to robotics course be relevant to employers if I want to get a job in the automation/robotics field in the future? How useful/relevant would it be to them?

I'm assuming the Machine Learning class is important but maybe not? I'm not really a programmer(at least in the sense of standard developer jobs) but I am starting to see job listings in signal processing want it. Same for the adaptive signal processing.

One issue is the adaptive signal processing class is offered less frequently but I will likely be graduating at the end of next year making it irrelevant.

What two do you think will be most useful? It seems like the Machine Learning class is maybe too important to leave out but I want to get others thoughts on this.​

 

[.Scott]

You're doing a Masters in EE with a focus in Signal Processing - so unless you thing that the Adaptive Signal Processing course is ground you have already covered, it would seem to be an obvious choice.

Machine Learning (ML) is very much Software Engineering - generally you don't tilt to an ML solution until other software approaches are deemed impractical or too expensive. So if you're not already up to making that decision, ML makes little sense. If you're not going to be a software engineer, there may be little value on the ML course. But not all ML courses are the same - in fact few are the same.

Even more than "ML", "Intro to Robots" would be an example of you-can't-judge-a-course-by-its-title.
Robotics is a combination of Mechanical Engineering, Electrical Engineering, and Software Engineering. All three are essential. So this would be a good candidate for your second choice.

 

[Qurks]

I agree it seems that way. A couple things though, 1. Most of the signal processing seems to use some form of machine learning(at least nearly all the research in my department does). Also I am seeing new job listings attempting to integrate machine learning into signal processing. I also had kind of an interest in computer vision and ML is relevant to that. A couple people in my EE department focus on Machine Learning.

Do you think ML would be largely useless to me when it comes to jobs? As in do you think as an EE graduate I wouldn't ever be considered for typical ML jobs in datascience/CV?

To be honest the signal-processing emphasis isn't super strong, most of the classes I have taken are relevant to both fields but I admit my programming background isn't very good since I have forgone most typical CS courses.

 

[jasonRF]

There is no absolutely right or wrong answer here. All you can do is figure out what you think you want to be doing in the near future, and take the classes that best prepare you to do that. The classes can have an influence over what jobs you can find your first few years out of school, but may not be so important after that.

By the way, I am an EE working in an R&D group in industry. We do quite a lot of signal processing, and while adaptive signal processing is bread-and-butter work for us, machine learning is certainly becoming much more important. There is no doubt that (at least in the short term) an understanding of machine learning can be a valuable asset if you want to work in signal processing.

I don't know anything about robotics so I won't pretend to offer advice on the value of your robotics course if you are interested in working in that field. It may be your most important course, or perhaps not...

Jason

 

[Jody]

Maybe compare the module descriptions with job descriptions? Are there reviews of the modules or the professor? Shoot... I'd go as far as looking at the textbook they might use and look at its reviews too.

 

[Qurks]

There is no absolutely right or wrong answer here. All you can do is figure out what you think you want to be doing in the near future, and take the classes that best prepare you to do that. The classes can have an influence over what jobs you can find your first few years out of school, but may not be so important after that.

By the way, I am an EE working in an R&D group in industry. We do quite a lot of signal processing, and while adaptive signal processing is bread-and-butter work for us, machine learning is certainly becoming much more important. There is no doubt that (at least in the short term) an understanding of machine learning can be a valuable asset if you want to work in signal processing.

I don't know anything about robotics so I won't pretend to offer advice on the value of your robotics course if you are interested in working in that field. It may be your most important course, or perhaps not...

Jason

Interesting, I don't do circuit design but I know some electromagnetics and a little bit of RF(In addition to my signal processing background). Would this be a problem for someone wanting to get a position in your company in signal processing?

 

[hagerww]

I'd agree with the other advice you've received... you should decide which classes give you the best tools you want for your desired path. However, if I were in your shoes - I would skip the Intro to Robotics course and, instead, volunteer as a mentor to one of many high school robotics programs. There are a number of different programs, Battle Bots is probably the most famous. I help a local FIRST team, and mentoring students through a 6-week design challenge would cover all the systems design and hardware/software interactions that I would expect to see in a graduate level Intro to Robotics course. This would also show (on a resume) as volunteer effort related to your field and hands on experience. If you go to the FIRST website, you should be able to search for teams in your area. I presume the Battle Bots and Vex websites, two other robotics programs, would have similar functionality.

 

[Qurks]

I'd agree with the other advice you've received... you should decide which classes give you the best tools you want for your desired path. However, if I were in your shoes - I would skip the Intro to Robotics course and, instead, volunteer as a mentor to one of many high school robotics programs. There are a number of different programs, Battle Bots is probably the most famous. I help a local FIRST team, and mentoring students through a 6-week design challenge would cover all the systems design and hardware/software interactions that I would expect to see in a graduate level Intro to Robotics course. This would also show (on a resume) as volunteer effort related to your field and hands on experience. If you go to the FIRST website, you should be able to search for teams in your area. I presume the Battle Bots and Vex websites, two other robotics programs, would have similar functionality.

Hi thanks,

One thing that complicates it is that I am considering a second masters degree in robotics but it requires the intro class. I think robotics will be a booming field in the future so a degree in it couldn't hurt right? Are dual master degrees looked down on?

Hard decisions, I am going to have to choose what path to take and I don't like it(lol). I like areas of EE but I don't really care for circuit design, at least at the semiconductor level.