AutomaticClassification

Code base for the study "Automatic classification of neurons by their morphology"

Url to the thesis

http://kth.diva-portal.org/smash/record.jsf?pid=diva2%3A1214317&dswid=-2523

Code

getData.py

Gathers the data from neuromorpho.org

transformData.py

Transform the gathered data so that is suites our classifiers.

classifier_assessment.py

Creates the classifiers using sklearn, runs the tests and displays the results.

cell_type_distribution.py

Plots the cell type distribution.

cell_type_accuracy_distribution.py

Plots the cell type accuracy distribution.

Data

after1000runs.json

This file contains the result from 1000 runs on each of the classifiers.

For each classifier each type is stored together with the list of the amount of correctly classified types and of the amount of missclassified types. E.g.

{'Purkinje': [11160, 3441]}

Images

The images used for the report.