Название: Bio-inspired Algorithms in Machine Learning and Deep Learning for Disease DetectionАвтор: Balasubramaniam S, Seifedine Kadry, Manoj Kumar TK, K. Satheesh Kumar
Издательство: CRC Press
Год: 2025
Страниц: 263
Язык: английский
Формат: pdf (true), epub
Размер: 29.5 MB
Currently, Computational Intelligence approaches are utilised in various science and engineering applications to analyse information, make decisions, and achieve optimisation goals. Over the past few decades, various techniques and algorithms have been created in disciplines such as genetic algorithms, artificial neural networks, evolutionary algorithms, and fuzzy algorithms. In the coming years, intelligent optimisation algorithms are anticipated to become more efficient in addressing various issues in engineering, scientific, medical, space, and artificial satellite fields, particularly in early disease diagnosis. A metaheuristic in Computer Science is designed to discover optimisation algorithms capable of solving intricate issues. Metaheuristics are optimisation algorithms that mimic biological behaviours of animals or birds and are utilised to discover the best solution for a certain problem. A metaheuristic is an advanced approach used by heuristics to tackle intricate optimisation problems. A metaheuristic in mathematical programming is a method that seeks a solution to an optimisation problem. Metaheuristics utilise a heuristic function to assist in the search process. Heuristic search can be categorised as blind search or informed search. Metaheuristic optimisation algorithms are gaining popularity in various applications due to their simplicity, independence from data trends, ability to find optimal solutions, and versatility across different fields.
Recently, many nature-inspired computation algorithms have been utilised to diagnose people with different diseases. Nature-inspired methodologies are now widely utilised across several fields for tasks such as data analysis, decision-making, and optimisation. Techniques inspired by nature are categorised as either biology-based or natural phenomena-based. Bioinspired computing encompasses various topics in computer science, mathematics, and biology in recent years. Bio-inspired computer optimisation algorithms are a developing method that utilises concepts and inspiration from biological development to create new and resilient competitive strategies. Bio-inspired optimisation algorithms have gained recognition in machine learning and deep learning for solving complicated issues in science and engineering. Utilising BIAs learning methods with machine learning and deep learning shows great promise for accurately classifying medical conditions.
Integrating bio-inspired algorithms with Machine Learning (ML) and Deep Learning (DL) models enhances Computational Intelligence (CI). These algorithms, like neural networks modeling the human brain, ant colony optimization, and particle swarm optimization, offer robust, efficient, and flexible models. Their inherent parallelism, adaptability, and self-organization capabilities significantly improve ML/DL model design, accuracy, and generalizability. Genetic algorithms optimize neural networks and hyperparameters, while swarm intelligence identifies optimal solutions, aiding DL model training. Additionally, bio-inspired algorithms enhance computing efficiency by finding near-optimal solutions with minimal computational cost, making them ideal for large-scale data processing. They excel in noisy, uncertain environments, maintaining performance under adverse conditions. Their adaptability allows ML/DL models to dynamically adjust to evolving data and problems. Bio-inspired algorithms also support localized, distributed ML/DL applications, promoting scalability and fault tolerance. Their interdisciplinary nature fosters innovation at the biology-computer science intersection, expanding ML/DL frontiers and addressing complex challenges in various fields.
This book explores the historical development of bio-inspired algorithms and their application in Machine Learning and Deep Learning models for disease diagnosis, including COVID-19, heart diseases, cancer, diabetes and some other diseases. It discusses the advantages of using bio-inspired algorithms in disease diagnosis and concludes with research directions and future prospects in this field.
Скачать Bio-inspired Algorithms in Machine Learning and Deep Learning for Disease Detection