Gxav_Sol_ASAP_round2

0. Unzip "Short_Answer_Gxav.zip". Read "Gxav Description.docx" to get familiar with the approach
1. Install R 2.15.1 (2012-06-22) -- "Roasted Marshmallows". Note previous versions could produce slightly different results because of some changes in the mgcv package
2. Open R
3. Store external data files, training set and test set files in a directory
4. Select as a R working directory the directory where the files were stored
5. Ensure all packages are installed 5a. SOAR # to manage memory usage 5b. Hmisc and MASS # for descriptive analysis 5c. Matrix # to handle matrix
   5d. RTextTools # to stem words 5e. tau # to produce n-grams 5f. tm # to remove duplicated blanks and punctuation marks 5g. clim.pact # to convert to lower case 5h. glmnet # to train a generalized linear model via penalized maximum likelihood 5i. glmnetcr # to train a multinomial generalized linear model via penalized maximum likelihood 5j. e1071 # to train Support Vector Machine 5k. gbm # to train Gradient Boosting Machine 5l. randomForest # to train Random Forest 5m. mgcv # to train Generalized Additive Models
6. Execute Gxav TRAINING CODE.R
7. Execute Gxav TEST CODE.R
8. 1 csv file should be generated “Gxav_Sol.csv” equal to the solution selected for the private leaderboard of the competition