Get started
Work

Parkinson’s Disease Identification using Reduced Kernel Classification

Year: 2025
Type: article
Abstract: Background: Artificial intelligence (AI) has begun to play an increasingly important role in the growing use of technology in several sectors. One of the areas where AI can greatly benefit is the medical sector. For instance, several hazardous diseases can be diagnosed early by applying neural networks, which are trained on an extensive array of populations. A good instance in which such applications are useful is Parkinson’s disease, due to the evolving nature of its symptoms. Materials and Methods: The study discusses the extreme learning machine (ELM), which is a specific kind of neural network with a feed-forward having a single layer. The kernel version of ELM and the reduced kernel approach have been employed; in the reduced kernel approach, the ELM is trained on a segment of the population. Results: The reduced kernel technique has been used for ELM with the objective of detecting PD. The sampling factor’s role and the kind of kernel were tested to examine their impact on the classifier’s performance. It was discovered that the value of the sampling factor in the range from 0.7 to 0.9 provided the best precision for detection with a kernel of type RBF. Moreover, the study was evaluated against the other high-tech methods based on SVM classifier and it was found that RKELM achieves 100% precision compared to SVM precision of about 91%–92%. Conclusion: The ELM approach brings forth impressive results compared to the traditional support vector machine (SVM) approach. After applying the EVM approach on various kernels and with various figures of samples from the overall population, the kernel has displayed 100% accuracy.
DOI: 10.4103/mjbl.mjbl_376_24
Institution: University of Misan
Language: English
Open Access status: gold
Estimated APC paid: $200
OpenAlex ID: W7125595300
FWCI:
Cites: 12
Cited by:
Related to: