Thoughts on Imperial College London's Masters in Machine Learning...
June 15, 2022 5:47 AM   Subscribe

... and Data Science? How well-regarded is this degree? How difficult?

I'm a 49-year-old American female with a background in programming but no work history the last 3 years due to health and family issues. I'm seriously considering getting a masters, probably in ML or DS. It must be 100% online and not require recent work or school references.

Ideally I'd be able to re-enter the workforce easily* after completion, and earn good money for the remainder of my work life so I can avoid living in a cardboard box in my old age.

How does this particular program stack up? Are there other online masters that I could be accepted to that are better?

* Despite my background in programming, hard work, and value I've provided to the companies I've worked for, I have been consistently undervalued and overlooked both at work and in the job search process, looked at as junior because of work gaps, etc. I know some of it is sexism, maybe some is due to my weight or introverted and non-showy personality, who knows what else. I say all this because "just get into ML/DS without a higher degree" is probably not something I will be allowed to do because 9f the way I am perceived.
posted by Flock of Cynthiabirds to Education (7 answers total) 2 users marked this as a favorite
 
I can't tell you about the course itself, but Imperial is one of the best universities in the world and its focus is on science and technology

Even without the rankings though, Imperial is regarded here in the UK as one of the finest universities in the country, for what that's worth to you.
posted by underclocked at 6:27 AM on June 15, 2022 [6 favorites]


I don't know anything about this program in particular, but in case it's something you need to hear, Imperial College itself is very highly regarded. For STEM it's generally ranked third in the UK (after Oxford and Cambridge) and in the top 10-20 globally. An Imperial degree carries a lot of weight.
posted by ManyLeggedCreature at 6:27 AM on June 15, 2022 [3 favorites]


Imperial College is very highly regarded in the UK (they won this year's University Challenge for a record fourth time). Edited for error.
posted by humph at 8:16 AM on June 15, 2022 [1 favorite]


Best answer: I am someone who in the past has hired people for data science roles and my own academic background is in computational astrophysics (although I have never worked in machine learning professionally and the people I hired were to do exploratory business data analysis and not ML). If I saw this on a CV I would definitely think: "this person is qualified for a wide range of roles requiring knowledge of data science and machine learning, they have a serious mathematical grounding in techniques".

That is a lot of roles out there right now.

The machine learning curriculum also looks substantial and has an appropriate focus on mathematical basis of the techniques. A lot of other ML masters courses are missing that and it limits the careers of their graduates to the business intelligence / marketing data track since they can't interpret new papers and have trouble learning and implementing techniques they weren't explicitly taught.

Disclaimer: I went to Imperial myself. It is very highly regarded generally in the UK and is well known to scientific employers in the US, I do not know if it is as well known by people who do not have a mathematical background themselves. I.e. someone with a PhD in statistics or machine learning would know and rate it highly but if the person hiring you doesn't come from that world themselves it may not have the name recognition in the US.
posted by atrazine at 8:25 AM on June 15, 2022 [6 favorites]


I went to a university for undergrad and grad school in the UK, and now work as a machine learning engineer in the US.

I'm going to answer this from the POV of someone who reviews CVs, i.e. assuming your goal is to get a job, rather than to get the best education possible.

First, I don't know anything about this course (or indeed any particular taught masters other than the incredibly famous and long lived non-remote ones like Cambridge Maths Part III). It's very unlikely that any random person who reviews your resume will know enough about a specific taught remote masters to have an opinion about its merits.

Second, while I think very few people will be familiar with this specific course, anyone with a technical background in the UK will pattern match on and be impressed by Imperial on your CV. It has a very strong reputation in science generally (top 3 in the UK, along with Oxbridge), and a strong reputation in computer science and machine learning (top 5ish along with Oxbridge and Edinburgh). It's less well known outside the UK, but it's not at all unknown. So, to the extent the reputation of the institution matters to you and the places you apply, this is a good choice. If you're planning to stay in the US, a top 10 domestic program (Stanford, MIT, the Ivy League, etc.) might be a better choice than a top 3-5 program overseas if you're optimizing solely for resume prestige. But Imperial is great

Finally, I can totally appreciate the value of a credential for someone in your position, but: many universities (even relatively solvent ones like Imperial) treat taught masters courses as profit centers, which makes teaching them obligations for faculty, rather than work they relish and do well. I don't know if that's a problem at this course or this university, but it's something to watch out for.
posted by caek at 10:12 AM on June 15, 2022 [4 favorites]


Hi. Background in chemoinformatics here, adjacent to bioinformatics. Previously worked at Imperial College. The course director is a good chap, but I suppose my recommendation is to think about which application of ML you specifically want to work in. No recent work history doesn't have to be a disadvantage if you have theoretical knowledge (such as a background in forensic science, or molecular biophysics, or whatever is the particular question the ML approach is looking to answer). The general attitude when I've worked in this field has been, it's easier to teach a scientist how to program, than to teach a programmer the science.
posted by Ardnamurchan at 11:03 AM on June 15, 2022


For your consideration-- A US based program will probably have stronger connections to industry in the US than international programs.
posted by oceano at 9:07 PM on June 15, 2022


« Older Tweens from LA to San Fran   |   Snowflake Online Workout Videos Wanted Newer »
This thread is closed to new comments.