Top-down learning path: Machine Learning for Software Engineers • Hacker News Board

Re: Top-down learning path: Machine Learning for Software Engineers

#11

Post by zuzoovn » Tue, Dec 27, 2016, 7:25 PM UTC

notheguyouthink wrote: post #7

This looks excellent, thank you for this! I actually get a chance to be involved with some very very basic/light ML at work, but i'm also completely new to the field. I really appreciate this style of guide, as i can hopefully jumpstart my knowledge, while i work to gain the required math knowledge. The Math will be my most difficult challenge, and i'm starting to brush up / learn courses to hopefully enable the prop…

There still seems to be a lot of people out there who believe they can never become proficient at mathematics. This really holds a lot of people back and it shouldn’t.

Once you have the motivation to learn something, because it solves an immediate problem you have, you will learn it.

Re: Top-down learning path: Machine Learning for Software Engineers

#12

Post by adamnemecek » Tue, Dec 27, 2016, 7:31 PM UTC

visarga wrote: post #9

It's great, but it's a mess - a heap of articles, github repos, books and such. A new learner can't deal with that. What a new learner needs is an interactive system to guide a learner through the process. Not a static list - one size doesn't fit all. I might know Python and not know math, another one might come from stats and Matlab but not know Python. One might know about SVMs and Naive Bayes but not about neural…

Check out Metacademy

Re: Top-down learning path: Machine Learning for Software Engineers

#13

Post by Eridrus » Tue, Dec 27, 2016, 8:03 PM UTC

mysteryme wrote: post #4

I appreciate a list of ML learning resources, but does anything about this strike you as strange? Over 9000 github stars and 1,260 forks for a repo started on 9th October????

GitHub has no concept of bookmarks, so people star repos instead.

Re: Top-down learning path: Machine Learning for Software Engineers

#14

Post by lobut » Tue, Dec 27, 2016, 10:02 PM UTC

zuzoovn wrote: post #10

visarga wrote: post #9

It's great, but it's a mess - a heap of articles, github repos, books and such. A new learner can't deal with that. What a new learner needs is an interactive system to guide a learner through the process. Not a static list - one size doesn't fit all. I might know Python and not know math, another one might come from stats and Matlab but not know Python. One might know about SVMs and Naive Bayes but not about neural…

It's a step-by-step guide on how to learning machine learning from overview to detail. Any comments are welcome. If you want the interactive system, please check "A Visual Introduction to Machine Learning: http://www.r2d3.us/visual-intro-to-machine-learning-part-1/"

I can easily infer your URL, but I think it includes an extra slash and double-quote.

Re: Top-down learning path: Machine Learning for Software Engineers

#15

Post by devilsavocado » Tue, Dec 27, 2016, 10:06 PM UTC

This is a great resource. I'd add Practical Deep Learning for Coders [0] to the video series list.

[0] http://course.fast.ai/index.html

Re: Top-down learning path: Machine Learning for Software Engineers

#16

Post by zitterbewegung » Tue, Dec 27, 2016, 10:08 PM UTC

visarga wrote: post #9

It's great, but it's a mess - a heap of articles, github repos, books and such. A new learner can't deal with that. What a new learner needs is an interactive system to guide a learner through the process. Not a static list - one size doesn't fit all. I might know Python and not know math, another one might come from stats and Matlab but not know Python. One might know about SVMs and Naive Bayes but not about neural…

I use http://dataquest.io and it works pretty well.

Re: Top-down learning path: Machine Learning for Software Engineers

#17

Post by notheguyouthink » Tue, Dec 27, 2016, 10:13 PM UTC

zuzoovn wrote: post #11

notheguyouthink wrote: post #7

This looks excellent, thank you for this! I actually get a chance to be involved with some very very basic/light ML at work, but i'm also completely new to the field. I really appreciate this style of guide, as i can hopefully jumpstart my knowledge, while i work to gain the required math knowledge. The Math will be my most difficult challenge, and i'm starting to brush up / learn courses to hopefully enable the prop…

There still seems to be a lot of people out there who believe they can never become proficient at mathematics. This really holds a lot of people back and it shouldn’t. Once you have the motivation to learn something, because it solves an immediate problem you have, you will learn it.

I don't think i can't be, i just know i have a lot to learn. My understanding of math is criminally small (despite enjoying the subject). Furthermore, ML seems to be on the advanced side of things.. at least, for those like myself.

I'm going to put in the work to learn it proper, but i suspect it will take me a fair bit of time. I've got many years to make up for.

Luckily there are many institutions online to aid in fundamental education like this :)

Re: Top-down learning path: Machine Learning for Software Engineers

#18

Post by ChristianGeek » Tue, Dec 27, 2016, 10:28 PM UTC

zitterbewegung wrote: post #16

visarga wrote: post #9

It's great, but it's a mess - a heap of articles, github repos, books and such. A new learner can't deal with that. What a new learner needs is an interactive system to guide a learner through the process. Not a static list - one size doesn't fit all. I might know Python and not know math, another one might come from stats and Matlab but not know Python. One might know about SVMs and Naive Bayes but not about neural…

I use http://dataquest.io and it works pretty well.

It's also $30-50/month.

Re: Top-down learning path: Machine Learning for Software Engineers

#19

Post by emilsedgh » Tue, Dec 27, 2016, 11:04 PM UTC

zuzoovn wrote: post #10

visarga wrote: post #9

It's great, but it's a mess - a heap of articles, github repos, books and such. A new learner can't deal with that. What a new learner needs is an interactive system to guide a learner through the process. Not a static list - one size doesn't fit all. I might know Python and not know math, another one might come from stats and Matlab but not know Python. One might know about SVMs and Naive Bayes but not about neural…

It's a step-by-step guide on how to learning machine learning from overview to detail. Any comments are welcome. If you want the interactive system, please check "A Visual Introduction to Machine Learning: http://www.r2d3.us/visual-intro-to-machine-learning-part-1/"

It was absolutely amazing! I suggest everyone who doesn't really know a thing about machine learning to take a look at this.

Re: Top-down learning path: Machine Learning for Software Engineers

#20

Post by amelius » Tue, Dec 27, 2016, 11:23 PM UTC

General ML question: is there some tool, supposedly based on ML, that can learn from my editing behavior, and automatically deduce and apply editing commands for repetitive editing tasks? This would be great e.g. when refactoring code.