Kaggle Greek Media Monitoring Multilabel Classification (WISE 2014)
install.packages("SparseM")
install.packages("caret")
install.packages("e1071")
install.packages("doParallel")
install.packages("foreach")
Download Training and Test data
Folder structure
KAGGLE-WISE2014
model_combine.R
writeAnswer.R - Create a submission file
score.r - Calculate F1 score
processLabelBySVM_linear.R
processLabelBySVM_radial.R
wise2014_svm_V01.R
wise2014_svm_V02.R
wise2014_svm_V03.R
wise2014_svm_V04.R
KAGGLE-WISE2014\Data
wise2014-test.libsvm.txt
wise2014-train.libsvm.txt
KAGGLE-WISE2014\Answer
sampleSubmission.csv
KAGGLE-WISE2014\Model
Set any parameters at the top of the program.
Main program is
wise2014_svm_02.R - SVM Linear
wise2014_svm_03.R - SVM Linear Parallel
wise2014_svm_04.R - SVM Radial Parallel
model_combine.R - Exploratory Analysis on a single label. investigate combining(stacking) SVM Linear and SVM Radial