HN Explorer

Ask HN: I'm getting a new job that involves machine learning. How do I kick ass?

15 points by ajushi ↗ HN
Hi guys,

I've been a PHP web dev all my career and I got this new job that involves a whole lot of machine learning. They are using Java too which I'm very familiar but haven't done extensive programming in yet.

I want to kick ass at this job. I want to do great work and be proud of it instead of going home everyday feeling mediocre.

Please help me do good at my work. What advice/suggestions can you give me? Any help would be greatly appreciated. Thank you so much in advance!

13 comments

[ 5.3 ms ] story [ 64.8 ms ] thread
Toshio
Hi. Congrats on your new job!

As far as advice/suggestions, please consider sharing a few more specifics, so the community may better understand where you're coming from.

ajushi
Thank you Toshio!

I'll be data mining about the usage of the millions of downloadable applications online.

jclos
I would advise you to get started with these books, which are practice-oriented (rather than theory-oriented such as the Mitchell book): http://www.cs.waikato.ac.nz/ml/weka/book.html with the Weka toolkit and/or http://www.liaad.up.pt/~ltorgo/DataMiningWithR/ with the R language and/or or http://shop.oreilly.com/product/9780596529321.do with Python

As for the theory, someone made a nice review of 10 popular ML books here http://zinkov.com/posts/2012-10-04-ml-book-reviews/ and http://www.stat.cmu.edu/~larry/all-of-statistics/index.html is a nice book on inferential statistics.

e-dard
Hi, machine learning PhD here - one way you can start is by brushing up on some fundamentals. A book such as Machine Learning, by Thom M. Mitchell is a reasonable start.

Also, in terms of applying ML, you could scoot through Andrew Ng's Machine Learning Coursera course (not sure if it's running at the moment, though).

Finally... One tip – typically I have always found that when you want to start apply ML to real-world problemss, start simple and only iterate when the results of your approach are not satisficing. This is usually because all the bleeding edge ML research/techniques don't consider a shit-load of real-world issues, like scaleability, applicability to wide-range of problem, unstructured or noisy data and so on.

jordanthoms
It's funny to see you mention that book, I literally have it next to me right now, studying for my exam next week :)
pav3l
+1 for Tom Mitchell's book. Although it may appear a little outdated, it is an excellent introduction to ML. Highly recommend doing the exercises (in whatever language you'll be working, not necessarily in C), especially the Neural Nets face recognition one. The book doesn't cover SVM, so you might want to learn about those elsewhere. Also if you don't have a good background in basic applied stats (linear models, logistic regression, etc), I suggest you brush up on that as well.
ajushi
Wow thanks for the suggestion. I'll definitely check it out.
pknerd
It has 2 much maths
S4M
How about checking the videos from the machine learning class on coursera by Andrew Ng?
pknerd
I wonder how did they hire you when you had no prior experience? You're lucky. I also want such job.
abhijat
Exactly what I would like to know too. I have been trying for a while now with not much luck, although I have picked up a decent amount of ML via self study.