The finest resources for Machine Learning & Deep Learning (coursera, fast.ai, kaggle and many more )

in programming •  7 years ago  (edited)

Its 2018 and if you want to get in on this right?!
Your first step is to study AI on a basic level to bring up some of your own nasty AI solutions.

1. Free online videos

Some excellent playlists to watch (ordered for better understanding):

Learn Python for Data Science



Intro to Tensorflow


Intro to Deep Learning (Udacity Nanodegree)


The Math of Intelligence


2. Online courses

Second you have to suck in knowledge from online courses, some excellent resources are:

Andrew Ng - Deep Learning Specialization

https://www.coursera.org/specializations/deep-learning

About Andrew Ng: Co-founder, Coursera; Adjunct Professor, Stanford University; formerly head of Baidu AI Group/Google Brain and really an "AI rockstar" in the AI community

Screen Shot 2018-02-26 at 16.40.33.png



Jeremy Howard - fast.ai

http://course.fast.ai/

Screen Shot 2018-02-26 at 16.41.31.png



3. Books

Third, read some books! One great choice would be this one:

Ian Goodfellow and Yoshua Bengio and Aaron Courville - Deep Learning

It helps you to with the terms in math of deep learning.
Available for free on http://www.deeplearningbook.org/



4. Collect your Data

Ever heard about "garbage in, garbage out"?
The quality of your data is the most important part of the machine learning pipeline even more so than the architecture of your AI model.

Github

The easy way to get data is to search for public datasets. You found awesome datasets on Github.
https://github.com/awesomedata/awesome-public-datasets

Screen Shot 2018-02-26 at 17.15.24.png



UC Irvine Machine Learning Repository

https://archive.ics.uci.edu/ml/index.php

Screen Shot 2018-02-26 at 17.16.25.png



Kaggle Datasets

https://www.kaggle.com/datasets

Screen Shot 2018-02-26 at 17.18.00.png



Create/buy your own datasets

DataCircle - where you can buy or directly exchange datasets with other people
https://datacircle.io/

Screen Shot 2018-02-26 at 17.23.27.png



Scrape data yourself using python with a library like Scrapy, Selenium, Beautiful Soup 4 etc.
Example: Retrieve Images from Wikipedia - https://gist.github.com/iwek/3100809

Screen Shot 2018-02-26 at 17.29.58.png

Diggernaut - turn website content into datasets. No programming skills required
https://www.diggernaut.com/

Screen Shot 2018-02-26 at 17.31.19.png



Last words

Keep in mind, big companies like Google have a huge advantage when it comes to building horizontal products that can apply to many industries like image recognition, language translation or infrastructure. But the advantage that smaller players have, like you, is that you can move fast on a single problem vertically. The big ones dont have to time to tackle every single niche problem. But you do! You can focus on the enterprise and build some niche solution that would help companies (and sell it to them).

One way to raise awareness of your product is to raising your own personal profile, establish yourself as an AI thought leader and publish blog posts on https://medium.com . Create blog content that answers fundamental questions about AI and build an audience. Then share it on Social Media or on Hacker News (https://news.ycombinator.com).
Example https://medium.com/@dtfoster



Screen Shot 2018-02-26 at 17.13.08.png



Let me end this with an inspirational meme :

Screen Shot 2018-02-26 at 17.11.06.png



If you enjoyed this article please Upvote and Resteem :)



Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Greetings @tai-euler! Your post was chosen at random and was resteemed as part of Shareables' campaign.

@Shareables, we resteem anything we find shareable. That means good quality content!

By upvoting this notification, you're supporting our campaign in Steemit. For more information about our campaign, click here!

God bless from us @Shareables!

well, thank you!