Abstract
Biological data analysis is one of the most important and challenging tasks in today's world. Automated analysis of these data is necessary for quick and accurate diagnosis. Intelligent computing-based solutions are highly required to reduce the human intervention as well as time. Artificial intelligence-based methods are frequently used to analyze and mine information from biological data. There are several machine learning-based tools available, using which powerful and intelligent automated systems can be developed. In general, the amount and volume of this kind of data is quite huge and demands sophisticated tools that can efficiently handle this data and produce results within reasonable time by extracting useful information from big data. In this chapter, the authors have made a comprehensive study about different computer-aided automated methods and tools to analyze the different types of biological data. Moreover, this chapter gives an insight about various types of biological data and their real-life applications.
TopIntroduction
In the field of diagnosis and medical research, analysis of biological data is inevitable. We can explore and extract some precious information from various types of biological data. To mine some knowledge from these data requires the involvement of the domain experts. Although humans are highly intelligent, there are some inherent problems associated with the human observers. For example, experiments done by the humans are subject to errors. It may also happen that, same data when diagnosed two times by the same observer may produce the different results. It is purely subjective and depends on the present state of the observer (Chakraborty, Chatterjee, et al., 2017a; Hore et al., 2015).
In general, biological sources generates a huge amount of data which is often considered as the big data (Prabha, Rai, & Singh, n.d.). Detailed study of these data is required for generating any useful information from the raw form. Now, humans are hardly capable of handling such a huge amount of data in a stipulated amount of time. This is an another problem associated with the human observers. Moreover, the analysis of these data is significant in the diagnostic field to provide proper treatment. The inaccuracies in the diagnostic process can be very costly in terms of the life of the patients.
Automated systems are highly beneficial in analyzing and mining useful information from the biological datasets. The term ‘data mining’ is a well-known technology which is used to find hidden information from large databases and big data. The exploration process is often known as ‘knowledge discovery’. Now a days, it is very hard to think the world without computers. The technical explosion in the automated systems makes the life easier. Automated systems are highly efficient and can mimic the diagnostic process as made by a human expert. Modern applications can effectively search the ocean of the data and discover useful information with high accuracy and in moderate time (Chakraborty, Chatterjee, Ashour, Mali, & Dey, 2017).
Most of the computer aided diagnostic systems are based on the artificial intelligence (AI) based tools. Artificial intelligence mimics the human intelligence and helps a machine to behave intelligently (“Artificial intelligence,” n.d.). In general, it can induce the cognitive nature in a machine. AI based tools provides the power to a computer to perform certain tasks very efficiently like humans (even better in some situation) (Boden, 1998).
In recent years, a huge growth in Artificial intelligence based technologies can be observed that can change the standard of the life. Artificial intelligence provides a way to make a machine learn. This technology is known as machine learning which is the blessing of artificial intelligence. Machine learning algorithms are used to make a machine learn and to avoid programming for each and every problem instance (Mitchell, 1997). These algorithms follows a set of rules for learning as well as for producing results. The performance of these algorithms are measured in terms of the accuracy and some other parameters. The machine learning based approaches can be broadly categorized as given follows:
Key Terms in this Chapter
Machine Learning: It is the ability of a system to learn or adapt something automatically from the environment, that is, experiments performed or the data being shown to the system and can make some decision in the unknown environment without any human intervention.
Artificial Intelligence: It is the ability of the machines to mimic the human intelligence and act accordingly.
Biomedical Image Analysis: Study of the biomedical images of various modalities using digital image processing techniques to detect and diagnose different diseases and help the medical investigation.
Data Mining: It is the process of determining hidden patterns from the dataset using various methods like machine learning, statistics, etc.
Intelligent Computing: It generally refers to the ability of a system to gather some knowledge from the data or the experiments.
Big Data Analysis: It is the process of analyzing huge amount of varied data to explore hidden relation from the dataset.
Biological Data Analysis: Study of the data which are acquired from the biological sources to interpret and find some hidden information.
Performance Evaluation: Evaluate the performance of any method or algorithm in terms of some parameters (need not to be always quantitative).