08.06.2015 Views

Building Machine Learning Systems with Python - Richert, Coelho

Building Machine Learning Systems with Python - Richert, Coelho

Building Machine Learning Systems with Python - Richert, Coelho

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

• <strong>Machine</strong>d <strong>Learning</strong>s at http://www.machinedlearnings.com<br />

Appendix<br />

° ° The average pace is one per month, providing more applied topics;<br />

often revolving around learning big data<br />

• FlowingData at http://flowingdata.com<br />

° ° The average pace is one per day, <strong>with</strong> the posts revolving more<br />

around statistics<br />

• Normal deviate at http://normaldeviate.wordpress.com<br />

° ° The average pace is one per month, covering theoretical discussions<br />

of practical problems. Although being more of a statistics blog, the<br />

posts often intersect <strong>with</strong> machine learning.<br />

• Simply statistics at http://simplystatistics.org<br />

° ° There are several posts per month, focusing on statistics and big data<br />

• Statistical Modeling, Causal Inference, and Social Science at http://<br />

andrewgelman.com<br />

° ° There is one post per day <strong>with</strong> often funny reads when the author<br />

points out flaws in popular media using statistics<br />

Data sources<br />

If you want to play around <strong>with</strong> algorithms, you can obtain many datasets from the<br />

<strong>Machine</strong> <strong>Learning</strong> Repository at University of California at Irvine (UCI). You can<br />

find it at http://archive.ics.uci.edu/ml.<br />

Getting competitive<br />

An excellent way to learn more about machine learning is by trying out a<br />

competition! Kaggle (http://www.kaggle.com) is a marketplace of ML competitions<br />

and has already been mentioned in the introduction. On the website, you will find<br />

several different competitions <strong>with</strong> different structures and often cash prizes.<br />

The supervised learning competitions almost always follow the following format:<br />

• You (and every other competitor) are given access to labeled training data<br />

and testing data (<strong>with</strong>out labels).<br />

• Your task is to submit predictions for the testing data.<br />

• When the competition closes, whoever has the best accuracy wins. The prizes<br />

range from glory to cash.<br />

[ 263 ]

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!