Predicting job salaries from ads - a Kaggle competition
Python Other
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
feature_selection
optional
regression_as_classification
.gitattributes
.gitignore
2vw.py
2vw_loc.py
LICENSE
README.md
add_dummy_salaries.py
first.py
split.py . Feb 20, 2013
unlog_predictions.r
update_locations.py
update_locations_fixed.py fixed a bug in update_locations.py Mar 23, 2013

README.md

Predicting advertised salaries

See http://fastml.com/predicting-advertised-salaries/ for description.

2vw.py - convert a combined train+test file to VW format
2vw_loc.py - the same, but for data transformed with update_locations.py
add_dummy_salaries.py - add dummy salaries columns (2) to a test file; drop headers
first.py - Take some lines from the input file and save them to the output file
split.py - split a file into two randomly, line by line
unlog_predictions.r - convert VW's log predictions back to a normal scale by taking exp()
update_locations.py - replace location columns from the original file with parsed location (five columns) - slightly buggy
update_locations_fixed.py - a fixed version of update_locations.py