Data Description
You are given three different data files which are train_LR, train_HR and test_LR. Both training files should be used to build your machine learning model. In the training sets, you are given low-resolution(LR) and high-resolution(HR) connectivity matrices of brain. First, you will use 5-fold cross-validation on the training set to design your model which predicts HR brain connectivity sample from an input LR sample. This will allow you to evaluate the generalizability and performance of the model. Once the model is fixed (no more changes in its composing components), you can retrain it using the whole train set, then test it on the independent test set to predict test_HR —which is the solution file that you are NOT given. You will submit your predictions through Kaggle, and your file will be evaluated and scored which will be shown in Leaderboard.
Once you obtain your predicted test_HR, melt it (see project PDF file) and upload it to the system.
File Descriptions
train_LR.csv: the training set for low-resolution matrices
train_HR.csv: the training set for high-resolution matrices
test_LR.csv: the test set for low-resolution matrices
sampleSubmission.csv: A sample submission file