Help me pick optional modules

What are your thoughts on neural computation and machine learning

You need a ton of mathematics (linear algebra, calc 2, 3, ordinary and partial diff. eq., above all numerical analysis, statistics) to be good at that.

surprised those are options then, because the course doesn't really have many math requirements.

It's weird how they put masters with introductory courses. What's your bachelor?

I agree for the mathematical background though. IT literally is a branch of mathematics after all, and that kind of studies makes the difference between a quick self-taught programmer, who can probably do some work with a classic high-level imperative language, such as Java or C#, and someone who can do research or understand systems and concepts that are more complex than on-click events in JavaScript : advanced functional programming, concurrent and distributed computations, graph theory and applications, etc.

Not spitting at web developers, that kind of jobs is needed too, but building an application in Haskell clearly shows a higher level in theoretical IT than one written in Java.

What I meant for the NISO course is that it can be great as an optional module (i.e. to complete the IT training), to better visualise the applications of some forms of algorithms, and learn that some systems seen in nature come from centuries of evolution, and are physically shown to be the best to solve a certain problem. It can also help build AIs for video games for example. It's a great mind-opener at the end of a curriculum. But yeah, if the compulsory modules are not a complete basis, better take more software engineering then.

Chemistry. Hated it so I wanted to do switch career path

Thirding this. SE is a bunch of bullshit, just get work experience; arguably if you've never programmed before an intro to SE might be vaguely useful but FSE will more than do that.
Out of the optional modules, stuff like crypto, ML or NN all look cool but they're stuff you'd want to do a full master's on. If you're curious, just download the slides if you can, and read about on like wikipedia or other online references/tutorials, you're likely to get easily as much if not more knowledge that way than a half-year module will give you.
Something like stats is always useful everywhere, and the good thing about it is while sure you could always find a textbook, I think it's a better module to ask questions on. For crypto/ML/etc. you can almost always find the answers online, but for maths stuff it's actually helpful to have professors to go to with questions.

They likely cover basic intro. So, stuff you can read on wikipedia. This is why I'm claiming that you'd really want to do a full course on them: only then will you actually learn interesting in-depth stuff you can't find in a few articles online.

Someone who did this course said this:

"Please don't choose the dosser route and go for the two Software Engineering modules! It's generally only taken by people wanting to put minimal effort in and is taught by an academic who hasn't been anywhere near the industry in 20 years and is teaching methodology that was old even then. "

>Zero mathematics and a single optional statistics course.
>This kind of people laugh at me in the economics threads spouting the "not a science" meme.

I did some math during my undergraduate. Don't remember much tho. Will i really need to know a lot of math where googling won't be sufficient?

Thing is, googling gives you math papers and algorithm explanations. You need a good background and intuition for that.
I think you'd make an excellent software engineer for having been a chemist. I suppose your mind naturally solves conditional combinatorial problems and compatibility resolutions. You'd probably make great interfaces and abstractions. You needn't concern yourself with too much math but in that case I recommend you pick up a good book on computer architecture like Structured Computer Organization by Tanenbaum. Don't have to read and understand every detail, just realise how and why processors work and that hardware suits software, not the other way around. Will help you.

Thanks

If you have no real formal CS education data structures & algos and intro to CS(assuming its not just some shit programming course) is where you should start.