So I am a B.Tech Graduate interested in machine learning .After looking through loads of online material I feel little overwhelmed . Is there a structured path to go in for machine learning.
There recently was a thread on Hacker News: https://news.ycombinator.com/item?id=18793849
I have struggled with this question for SOME TIME. I have played with MIT/OCW and others. The field is SO big now and there are a lot of resources but not put together for a methodical complete online program.
After a year of self study and meeting with peeps I think it is good to make sure you have a good background in:
- How is your linear algebra?
- How is your statistics? Linear regression, Covariance, Correlation, …
- Do you know SQL?
- How is your Python or R?
Early on I was given one paper as a starting point.
- Top 10 Data Mining Algorithms - http://126.96.36.199:802/algorithms/10Algorithms-08.pdf
- I was given this paper to start with… the paper is from 2008 but it contains the basics that you NEED to know first. A few things are obviously missing (Logistic regression) but this paper gives you a good overview. You need to learn SVM and Adaboost and Decision trees, clustering, etc.
- Many people told me NOT to start with Deep Learning. BC you need to know at least Logistic Regression first. I tend to agree.
Ping me if you want to chat,
Go for andrew NG’s courses. They are the best. hands down.
Hi! I just read your response to this thread and I find it really helpful, so thanks!
I’ve just finished a PhD in Engineering and I’m looking to switch a career into Data Science. I have very good knowledge of linear algebra, statistics and Python. I’m picking up SQL but the next part I wasn’t sure. So thanks for sharing this!
Sure thing, best of luck.
Looking over this list there is one thing I could have added.
This is interesting. I am making the switch from programming to data science and I look forward to exploring these resources also.