2019 Data Science Lab Project. Team: Casamento, Palumbo, Stolfo.
- data: folder containing all available data.
- Age discrepancy_t_tests: computes t-tests pvalues for the mean of age discrepancy in patients with 4 of the most frequent disorders versus healthy patients.
- AgeDiscrepancy_controllingForAge: computes pvalues for the mean of age discrepancy in patients with 4 of the most frequent disorders versus healthy patients, controlling for age.
- AgeHealthy: tests on quality of age prediction on healthy patients only (CDE versus XGBoost).
- Age_globalCorticalThickness: age prediction from single feature (global cortical thickness).
- Age_EEG: Age prediction from EEG data (average+cluster features). Every model is fine-tuned with grid search.
- Age_MRI: Age prediction from MRI data (structural only and structural+DTI). Every model is fine-tuned with grid search.
- Age_EEG_MRI: Age prediction from MRI and EEG combined. Every model is fine-tuned with grid search.
- CDE: age prediction with Conditional Density Estimation model.
- FindTestset: A random subset of 120 patients is selected as testset. The corresponding test indices are found in the 'data' folder, named 'test_IDS.csv'
- prediction_with_EEG: Disease prediction from EEG data
- SWAN_EEG: SWAN score prediction from EEG data
- SWAN_MRI: SWAN score prediction from MRI data
- WISC_EEG: WISC score prediction from EEG data
- WISC_MRI: WISC score prediction from MRI data
- utils.py: utils functions for datasets creation, cross validation and visualization.