rafadaguiar / ml_class

code for the ml class

Machine Learning

class taught by Hilary Mason

Install

• JSONView Chrome extension
• Python (2.5, 2.6 or 2.7). Note forom scipy.org: NumPy installer should be used with the Python from http://python.org, not with Apple Python. These two are indeed incompatible, for one the python.org version is 32-bit while Apple version is 64-bit. Apple is also way behind on security updates, so normally python.org is the way to go.
• NLTK
• NumPy
• pycluster
• hcluster

Verify your install (the commands below after '$' should be typed at the command prompt, the rest is sample output)

$ python --version
Python 2.6.1

any version of 2.5, 2.6 or 2.7 is fine. 3.0 is not going to work.

$ python
Python 2.6.1 (r261:67515, Jun 24 2010, 21:47:49) 
[GCC 4.2.1 (Apple Inc. build 5646)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import nltk
>>> 

If you don't get an error, it means it worked. Type exit() to leave the interactive python console.

References

• Peter Skomoroch's dataset Bookmarks
• Data Source Handbook by Pete Warden

Classifying Web Documents

Register for an API key at http://developer.nytimes.com/apps/register and select "Article Search API"

example command-line: curl "http://api.nytimes.com/svc/search/v1/article?query=jazz&api-key="

python nytimes_pull.py

creates two files "arts" and "sports"

Naive Bayes Clasifier

python