Machine Intelligence Techniques For Biomedical Data Analysis And Classification

Abstract

Throughout history, human health has been under constant threat from various diseases, newlineparticularly in our modern times, where changing lifestyles and environmental pollution newlinehave exacerbated the problem. Many of these diseases silently infiltrate our bodies, newlineleaving cryptic traces within our genetic makeup. Extensive investments in genetic newlineresearch have brought to light that the root cause of numerous illnesses lies in genetic newlinemutations. Identifying these mutations at an early stage holds the key to effectively newlinecuring or managing such diseases for a lifetime. newlineA crucial approach in this pursuit is the extraction of vital information from biomedical newlinedata, aiding medical experts in making informed decisions and developing innovative newlinedrugs. Its initial work presents three improvised machine learning classifiers wrapped newlinewith a recently developed evolutionary optimization algorithm to design robust newlineclassification models. These models not only significantly boost accuracy but also newlineenhance diagnostic speed and reliability. newlineAgain, three hybridisation models have been applied on three clinical datasets such as newlineIndian diabetes, Parkinson, and Liver disorder patient datasets. The contributions of the newlinethesis, towards clinical data classification are summarized below. newlineand#61623; In the first attempt, a novel optimization algorithm, namely, Fruit Fly newlineOptimization (FFO) is applied to optimize the basic Extreme Learning Machine newline(ELM) classifier which classifies the three small clinical datasets, such as Indian newlinediabetes, Parkinson, and Liver disorder patient datasets. newlineand#61623; In the second attempt, another variant of ELM, namely, OSELM classifier is newlineoptimized by Adaptive FFO (AFFO) algorithm. This model is tested for the same newlineclinical datasets, such as Indian diabetes, Parkinson, and Liver disorder patient newlinedatasets. newlineand#61623; In the last approach, a Levy flight-based FFO (LV-FFO) algorithm optimized newlineKernel Extreme Learning Machine (KELM) is applied for diagnosis of Indian newlinediabetes, Parkinson, and Liver disorder patient datasets. newlinevii newlineGene expression dat