Repository for analysis of schizophrenia dataset
This projects needs python 3 and the dependencies specified in environment.yml
.
To use the classical machine learning pipeline a configuration file needs to be created as a JSON file (details below).
To run the analysis execute: python analysis/classical_ml/classical_ml.py cfg_file_path
.
The configuration file describes how the analysis will be run. It should be on a JSON format with the following structure:
Attribute | Value | Remarks |
---|---|---|
read_mode | String that specifies the type of data to be used | Value should be one of the following: BIDS , table , h5 |
paths | JSON Object with specification of important paths | The paths that can be described here are:
|
images | JSON Object with the description for data loading and preprocessing | The possible keys (default value) are:
|
training | JSON Object with the description for data training | The possible keys (default value) are:
|
holdout | JSON Object with the description for inference in holdout set (optional) | The possible keys (default value) are:
|
model_selection |
It can have one of the following values:
|