Resting State Dynamics Meets Anatomical Structure: Temporal Multiple Kernel Learning (tMKL) Model
python3 sklearn
Open matlab and run the following:
run <tMKL_dir_path>/CODE/main_script.m
- CODE - contains scripts and functions
- main_script.m - main script for running tMKL
- get_train_struct.m - wFCs generation, vectorize wFCs, spectral embedding - projection to manifold, GMM - assign wFCs to states
- run_tMKL.m - run MKL model for every state
- grand_average_prediction.m - evaluate model performance
- INPUT_MATS - stores the data, including SC FC matrices and BOLD timeseries data.
- INTER_MATS - stores intermediate representations : wFCs, cluster assignments, transition matrices, tMKL model parameters etc.
- OUTPUT_MATS - stores final model performance results