1. How to predict on new data ?
a. Put gzipped csv file named input.csv.gz in newdata/ .
The format should be the same as validation set.
b. Start R in this directory.Run source('predict_new.R')
d. the output is newdata/ouput.csv
2. Environment
a. System
Linux lq-laptop 3.2.12-gentoo #1 SMP Tue Mar 27 21:09:25 CST 2012 x86_64 Intel(R) Core(TM) i7-2640M CPU @ 2.80GHz GenuineIntel GNU/Linux
b. R verison
2.15.0
c. R packages
"lars" "Matrix" "lattice" "RWeka" "openNLP" "Aspell" "tm"
3. Time of prediction
Time of predicting on validation set:
a. simple feature extraction : 39.68933s
b. prediction : 52.90542 s