Title : Leveraging artificial neural networks and harmony search algorithm for neurological data analysis
Abstract:
This study explores the application of Artificial Neural Networks (ANN) combined with the Harmony Search Algorithm (HSA) to enhance data analysis in neuroscience, particularly for detecting and classifying anomalies in neurological datasets. Data for this research were obtained from the UCI German Dataset, consisting of 1,000 samples, including 700 normal and 300 abnormal cases. While originally designed for financial fraud detection, the methodology demonstrates adaptability for handling complex and high-dimensional data, such as those encountered in neurological research. The proposed Neural Network Harmony Search (NNHS) system optimizes ANN parameters using HSA to uncover hidden patterns in the data. For evaluation, accuracy was chosen as the primary performance metric, with the system undergoing multiple iterations to ensure reliability. Over 10 iterations, NNHS achieved an average accuracy of 86%, outperforming traditional methods evaluated under similar conditions. This approach addresses common challenges in neurological data analysis, such as data imbalances and computational complexity, making it a valuable tool for improving diagnostic precision and research outcomes. The results highlight the potential of NNHS to contribute significantly to advancing patient care and bridging the gap between computational neuroscience and clinical applications.
Keywords: Artificial Neural Networks, Harmony Search Algorithm, Neurological Data Analysis, Anomaly Detection, Computational Neuroscience, Diagnostic Precision, UCI German Dataset.
