Anyone trying to get into machine learning? I have a only a simple

b yourself

What is your statistical background?

>>7919622
Just be yourself.

Don't spend time on that. Learn how to program. You said it yourself, you are a bad programmer.

You can't begin any type of classification or clustering without an initial dataset. So learn how to download stuff from the internet. Scrape, farm an API, anything. Then learn how to store your data in a schema that's not shit. Then move onto visualization.

>>7919780
masters in finance with a bunch of courses on regression/time series analysis. I'm no expert, but I understand enough to stay interested and build on what I know.

>>7919802
I'm going through code academy's python course since I don't know python and the later stuff goes beyond loops, arrays and classes that I know how to use. I do agree that downloading stuff through code is pretty sweet and should be in my toolset. I aim to use ML in games once I know enough. I have a few ideas I'd like to experiment with but of course having a means to support myself while learning all these things would be fantastic.

File: tensorflow.jpg (5KB, 230x200px) Image search: [Google]

5KB, 230x200px

>>7919622
Check out Hinton's coursera course and do all the programming assignments for cs231n.github.io, they were immensly helpful for me, coming from physics. Then move on to kaggle competitions. Daniel Nouris tutorial on facial keypoint detection with convnets is wonderful.

ML is very in demand currently, there are lots of well paying jobs(mostly sales forecasting, anomaly detection, recommender systems type stuff) but you probably won't land one without a relevant degree.

If you can actually write good code, contribute to some ML libraries on github. That's probably your best bet for eventually landing a job.

File: peternorvig.jpg (31KB, 185x239px) Image search: [Google]

31KB, 185x239px

machine learning is useless. Have you read the peter norvig's work?

File: peterNorvig.jpg (227KB, 630x557px) Image search: [Google]

227KB, 630x557px

>>7919622
why even make this thread? don't you realise that piotr nirvug proved machine learning was impossible?

>>7921700
SHITTY BAIT fuck off from sci

>>7921705
clearly you AI fanatics don't understand deep learning,

>>7921716
I bet they haven't read peter norvig's work

>>7921700
>>7921704
>>7921716
You don't understand the "Artificial Intelligence: A Modern Approach" meme, I would highly recommend that you lurk moar and read "Artificial Intelligence: A Modern Approach" by Russel anf Norvig.

>>7921695
Might check it out after andre Ng's. I found out about the NEAT algorithm and it just blew me away with how clever it is. I'm gonna try to make that work in a game today.