Parkinson’s Disease diagnosis classification using gait cycle similarities
The objectif is to do Feature Extraction from the orginal dataset.
Then I used the new generated dataset with the new features with different classification machine learning algorithms to predict parkinson's disease.
I used Similarity measures like 'Euclidean Distance' ,'Time-Wraped Edit Distance', 'Dynamic Time Wrapping', 'Edit Distance on Real sequence': to measure similarity between Timeseries Data.
The timeseries data represent the Force VGRF(Vertical Ground Reaction Force) of people while walking.
*Feature extraction with Similarity Measurement (on Timeseries data)
*Dealing with unbalanced data
*Using Unsupervised and supervised Learning
*evaluate models
Dataset source : https://physionet.org/content/gaitpdb/1.0.0/