This class approaches Machine Learning as software with an API.
Most ML classes focus on under-the-hood topics like Gradient Descent.
This class presents Machine Learning more as a 'black box'.
This class helps you connect your ML project to a variety of APIs available from repositories like MADlib, scikit-learn, and LIBSVM.
The data we feed to our ML projects will come from Yahoo finance as stock prices.
Stock price data is constantly changing, diverse, and interesting.
It makes excellent learning material.
The first class will deal with getting your learning environment setup.
I have great success with running my ML projects on Linux hosts.
You can run Linux hosts inside your laptop using virtualization software.
Also you can run them remotely in large data centers for pennies per hour.
I may be able to get you some free compute time from local cloud providers.
The aim of this class is to get you proficient at building a simple system which pulls in prices once a day, issues predictions, and then helps you visualize the accuracy of your project.
Once you are finished with that, you will be ready for a wide variety of Machine Learning projects.
Here is a checklist of 3 things you need to bring to the first class:
- Good quality Laptop (4GB of RAM, 50GB of free disk, minimum)
- Cent0S 6.5 ISO file on disk, DVD, or thumb-drive:
- VirtualBox software on disk, DVD, or thumb-drive:
If you have problems with the above checklist,
try this checklist:
- Wimpy laptop
- An account on one of these cloud providers:
- Amazon AWS
- Digital Ocean
- RedHat OpenShift
- Google Compute Engine
If you dont have a cloud account, bring a credit card and I/we will help you get one.
If you come to class with only an iphone or ipad, you will get a sub-optimal learning experience.
This class is for people who want to actually write ML code using a laptop.
If you want to interact with other students outside of class, join this forum using a gmail account: