lot of research has been done to predict Parkinson’s disease in a person,
however less work has been done to predict its severity in patients. These
works have used various machine learning techniques like random forests,
support vector machines, decision trees, dynamic neural network, and Hidden Markov Models.
For effectual diagnosis of the Parkinson’s Disease (PD), several types of
classification techniques were employed by Das2. The performance score of the
classifiers was computed based upon various evaluation methods. As per the
results of application scores, they found that Neural Networks (NNs) classifier
obtains the best result, giving 92.9% accuracy.
Genain et al.
3 have used Bagged decision trees (random forests) to predict continuous measures of Parkinson’s
Disease Severity from voice recordings of patients. They have used the
University of California Irvine Parkinsons Telemonitoring Dataset and dataset
from Synapse.org and achieved model accuracy to an RMSE of 2.
Akash P. Manwatkar et al. 4 have created a classifier that extracts features
from voice recordings of Parkinson’s disease patients and detects the severity
level of the disease. The dataset used was provided by Synapse.org, containing
voice recording of patients along with their demographic information as well as
their PDRS and Hoehn & Yahr scale score. Several voice filtering algorithms
such as cepstral mean normalization, band path filter, and spectral subtraction
were applied to pre-process the data and then combinations of machine learning
algorithms such as support vector machine and decision tree classified each
recording to the severity. The obtained result had an accuracy of 62.2% in
predicting Hoehn Yahr score and RMS error of 9.57 in predicting PDRS score.
the research work, ‘Multiclass classification of Parkinson’s disease using
different classifiers and LLBFS(Local Learning Based Feature Selection) feature
selection algorithm’, done by Ben Malek et al.5, a 40-features dataset has
been used, where after analyzing all the features, 9 were selected out of 40 to
classify PWP subjects into four classes, based on their unified Parkinson’s disease
Rating Scale (UPDRS). The various techniques
used for this work were Subspace discriminant, KNN and SVM which provided an
accuracy of 96.5%, 67.2% and 90.1% respectively.