I'm surprised this is on HN, probably the lack of downvote buttons. It's just banal and generic advice, there's zero useful information for any mildly experienced programmer there.
1. Don't put machine learning on a pedestal. – Do programmers really do this mistake? And what exactly does that mean?
2. Don't write machine learning code. – A classic programming principle, not specific to machine learning at all.
3. Don't do things manually. – Oh really? Thanks, I didn't know that.
4. Don't reinvent solutions to common problems. – Obviously, the same principle as 2.
5. Don't ignore the math. – Ok, good point, but if you want to get serious with ML, it's difficult to avoid maths.
Comments
I'm surprised this is on HN, probably the lack of downvote buttons. It's just banal and generic advice, there's zero useful information for any mildly experienced programmer there.
1. Don't put machine learning on a pedestal. – Do programmers really do this mistake? And what exactly does that mean?
2. Don't write machine learning code. – A classic programming principle, not specific to machine learning at all.
3. Don't do things manually. – Oh really? Thanks, I didn't know that.
4. Don't reinvent solutions to common problems. – Obviously, the same principle as 2.
5. Don't ignore the math. – Ok, good point, but if you want to get serious with ML, it's difficult to avoid maths.
Agree. I found this post by John Langford more informative, on a similar vein:
http://hunch.net/?p=2562
Points 2 and 4 literally means, "Stand on the shoulder of giants"!