Where to start Machine Learning?

Question

I've recently stumbled upon machine learning, generating user recommendations based on user data, generating text teaser based on an article. Although there are tons of framework that does this(Apache Mahout, Duine framework and more) I wanted to know the core principles, core algorithms of machine learning and in the future implement my own implementation of machine learning.

Where do I actually start learning Machine Learning as its basics or Machine Learning starting with concepts then to implementation? Although I have an weak to average math skills(I think that this will not hurt? If So what branches of mathematics should I study before Jumping to Machine Learning?)

Note that this is not related to my academics rather I want to learn this as an independent researcher and I am quite fascinated how machine learning works

The python scikit-learn library is a good thing to be aware of. I've heard the documentation is very good, and it will give you an overview of many standard topics in machine learning. — littleO, Sep 22 '13 at 09:08

score 8 · Answer 1 · answered Jun 20 '13 at 07:22

8

You can start machine learning from Coursera Machine Learning

answered Jun 20 '13 at 07:22

Mohan

14,856

do I need to know in depth statistics? – user962206 Jun 20 '13 at 07:45
No, not at all. You don't need any prior knowledge of statistics. Only basic linear algebra like matrix multiplication and some very elementary calculus. – Mohan Jun 20 '13 at 07:51
so I guess I need to re-learn my basic statistics , basic linear algebra and calculus? – user962206 Jun 20 '13 at 08:33
Yes. But there is a nice review of these materials in the course. – Mohan Jun 20 '13 at 09:09
Okay thanks! I'll just immediately start this course! thanks! – user962206 Jun 20 '13 at 11:16
Probabilistic Graphical Models is also a great course. – Steve P. Jun 23 '13 at 16:49

score 5 · Answer 2 · answered Jun 20 '13 at 07:37

5

Try to read Hastie, Tishirani, Friedman "Elements of Statistical Learning". It is a nice book. If you are specifically interested in unsupervised learning, then try the quite new Kogan "Introduction to Clustering Large and High Dimensional Data": algorithms are explicitly given.

answered Jun 20 '13 at 07:37

Avitus

14,018

1

If that Hastie et al book is too math heavy, (some of) the same authors have a simpler book, with R in its title. – kjetil b halvorsen May 19 '15 at 17:49
this is a very good comment: the simpler book with R examples is an excellent introduction to the subject. – Avitus May 20 '15 at 07:22

score 5 · Answer 3 · answered Sep 22 '13 at 08:47

5

If you need some ideas for practical projects to immerse yourself into machine learning, you might be interested in looking at Programming Collective Intelligence by Toby Segeran.

answered Sep 22 '13 at 08:47

Filip Dupanović

151

score 3 · Answer 4 · answered May 19 '15 at 17:42

3

Coursera Machine Learning is one of the best places to start, but there are countless resources. There's a huge, wonderful open list of links at Github. I definitely recommend you take a look at it.

There are also another great resources online, like those I list below:

In-depth introduction to machine learning in 15 hours of expert videos
Deep Learning Tutorial (@ ufldl.Stanford.edu/tutorial/, can't post the link because I'm out of mana, I mean, not enough reputation yet).

Hope it helps!

answered May 19 '15 at 17:42

Bruno Lucattelli

131

When you get enough rep you can come back and edit in the link! – kjetil b halvorsen May 19 '15 at 17:47
@kjetilbhalvorsen yeah, thanks! – Bruno Lucattelli May 19 '15 at 17:48

score 2 · Answer 5 · answered Jun 30 '17 at 02:20

In addition to great resources already mentioned, I highly recommend Kevin Murphy's book: "Machine Learning: A Probabilistic Perspective".

For implementations of machine learning algorithms, have a look at:

score 2 · Answer 6 · answered Jul 04 '18 at 08:56

2

Machine learning crash course from google is good place to start. Also as mentioned above, Machine Learning course from Coursera is also great material on machine Learning.

answered Jul 04 '18 at 08:56

darshanz

121

score 1 · Answer 7 · answered Feb 07 '19 at 07:43

1

I would also recommend the course Learning from data by Yaser Abu-Mostafa from Caltech. An excellent course!

answered Feb 07 '19 at 07:43

jjepsuomi

8,619

score 0 · Answer 8 · answered Jul 11 '22 at 05:54

If you really like to learn ML Algos in depth and from the scratch and want to try to implement them yourself, so you can learn and understand them better, I recommend Cornel University ML course.

For me this is thus far the best course I've ever seen on ML.

Where to start Machine Learning?

8 Answers8

Linked

Related