Название: Applications of Machine Learning and Deep Learning on Biological Data
Автор: Faheem Masoodi, Mohammad Quasim, Syed Nisar Hussain Bukhari
Издательство: CRC Press
Год: 2023
Страниц: 211
Язык: английский
Формат: pdf (true)
Размер: 10.18 MB

The automated learning of machines characterizes Machine Learning (ML). It focuses on making data-driven predictions using programmed algorithms. ML has several applications, including bioinformatics, which is a discipline of study and practice that deals with applying computational derivations to obtain biological data. It involves the collection, retrieval, storage, manipulation, and modeling of data for analysis or prediction made using customized software. Previously, comprehensive programming of bioinformatical algorithms was an extremely laborious task for such applications as predicting protein structures. Now, algorithms using ML and Deep Learning (DL) have increased the speed and efficacy of programming such algorithms.

Applications of Machine Learning and Deep Learning on Biological Data is an examination of applying ML and DL to such areas as proteomics, genomics, microarrays, text mining, and systems biology. The key objective is to cover ML applications to biological science problems, focusing on problems related to bioinformatics. The book looks at cutting-edge research topics and methodologies in ML applied to the rapidly advancing discipline of bioinformatics.

ML and DL applied to biological and neuroimaging data can open new frontiers for biomedical engineering, such as refining the understanding of complex diseases, including cancer and neurodegenerative and psychiatric disorders. Advances in this field could eventually lead to the development of precision medicine and automated diagnostic tools capable of tailoring medical treatments to individual lifestyles, variability, and the environment.

In bioinformatics, extracting knowledge from data using machine learning is a common practice. Predictions are made based on the best relevant model derived from the data. In several domains, including genomics, proteomics, and systems biology, algorithms (like Support Vector Machines (SVMs), random forests, Hidden Markov Models, Gaussian networks and Bayesian networks) have been implemented. But machine learning is not good at detecting features properly. To overcome this limitation, DL is introduced in the bioinformatics area. Disciplines such as image identification, Natural Language Processing (NLP), drug development, and bioinformatics have all benefited significantly from DL.

Highlights include:

Artificial Intelligence in treating and diagnosing schizophrenia
An analysis of ML’s and DL’s financial effect on healthcare
An XGBoost-based classification method for breast cancer classification
Using ML to predict squamous diseases
ML and DL applications in genomics and proteomics
Applying ML and DL to biological data