Development of high efficiency optimization algorithm based on new topology in particle swarm optimization for parkinson's disease

Parkinson's disease is a progressive neurodegenerative disorder, where new cells in the brain die slowly over time due to the loss of the substantia nigra causing the lack of dopamine. In order to tackle the accurate diagnosis of Parkinson disease issue, this research proposes to design of a new algorithm to efficiently discern Parkinson's disease patients from the healthy individuals using the binary particle swarm optimization method. The case-control study was designed to evaluate this algorithm by analyzing the electroencephalogram dataset of over 196 patients with Parkinson's disease and compare it with that of healthy controls of the same age and gender. The comparison includes other traditional deep learning, logistic regression, and binary particle swarm optimization techniques to find out the most accurate one between the three methods. In this thesis, for Parkinson's disease we used the co-evolution binary particle swarm optimization based on multiple inertia weight strategy. Also, the ant colony optimization method and the genetic algorithm has been used to reduce the features number from the dataset. For classification the support vector machine is used and compared with other methods. Also, for performance analysis the confusion matrix is used.