Soft computing

Last updated

Soft computing is an umbrella term used to describe types of algorithms that produce approximate solutions to unsolvable high-level problems in computer science. Typically, traditional hard-computing algorithms heavily rely on concrete data and mathematical models to produce solutions to problems. Soft computing was coined in the late 20th century. [1] During this period, revolutionary research in three fields greatly impacted soft computing. Fuzzy logic is a computational paradigm that entertains the uncertainties in data by using levels of truth rather than rigid 0s and 1s in binary. Next, neural networks which are computational models influenced by human brain functions. Finally, evolutionary computation is a term to describe groups of algorithm that mimic natural processes such as evolution and natural selection.

Contents

In the context of artificial intelligence and machine learning, soft computing provides tools to handle real-world uncertainties. Its methods supplement preexisting methods for better solutions. Today, the combination with artificial intelligence has led to hybrid intelligence systems that merge various computational algorithms. Expanding the applications of artificial intelligence, soft computing leads to robust solutions. Key points include tackling ambiguity, flexible learning, grasping intricate data, real-world applications, and ethical artificial intelligence. [2] [3]

History

The development of soft computing dates back to the late 20th century. In 1965, Lotfi Zadeh introduced fuzzy logic, which laid the mathematical groundwork for soft computing. Between the 1960s and 1970s, evolutionary computation, the development of genetic algorithms that mimicked biological processes, began to emerge. These models carved the path for models to start handling uncertainty. Although neural network research began in the 1940s and 1950s, there was a new demand for research in the 1980s. Researchers invested time to develop models for pattern recognition. Between the 1980s and 1990s, hybrid intelligence systems merged fuzzy logic, neural networks, and evolutionary computation that solved complicated problems quickly. From the 1990s to the present day, Models have been instrumental and affect multiple fields handling big data, including engineering, medicine, social sciences, and finance. [4] [5]

Computational techniques

Fuzzy logic

Fuzzy logic is an aspect of computing that handles approximate reasoning. Typically, binary logic allows computers to make decisions on true or false reasons (0s and 1s); however, introducing fuzzy logic allows systems to handle the unknowns between 0 and 1. [2] [6]

Unlike classical sets that allow members to be entirely within or out, fuzzy sets allow partial membership by incorporating "graduation" between sets. Fuzzy logic operations include negation, conjunction, and disjunction, which handle membership between data sets. [5]

Fuzzy rules are logical statements that map the correlation between input and output parameters. They set the rules needed to trace variable relationships linguistically, and they would not be possible without linguistic variables. Linguistic variables represent values typically not quantifiable, allowing uncertainties. [7]

Neural networks

Neural networks are computational models that attempt to mimic the structure and functioning of the human brain. While computers typically use binary logic to solve problems, neural networks attempt to provide solutions for complicated problems by enabling systems to think human-like, which is essential to soft computing. [8]

Neural networks revolve around perceptrons, which are artificial neurons structured in layers. Like the human brain, these interconnected nodes process information using complicated mathematical operations. [9]

Through training, the network handles input and output data streams and adjusts parameters according to the provided information. Neural networks help make soft computing extraordinarily flexible and capable of handling high-level problems.

In soft computing, neural networks aid in pattern recognition, predictive modeling, and data analysis. They are also used in image recognition, natural language processing, speech recognition, and systems. [3] [10]

Evolutionary computation

Evolutionary computation is a field in soft computing that uses the principles of natural selection and evolution to solve complicated problems. It promotes the discovery of diverse solutions within a solution space, encouraging near-perfect solutions. It finds satisfactory solutions by using computational models and types of evolutionary algorithms. Evolutionary computation consists of algorithms that mimic natural selection, such as genetic algorithms, genetic programming, evolution strategies and evolutionary programming. These algorithms use crossover, mutation, and selection. [11]

Crossover, or recombination, exchanges data between nodes to diversify data and handle more outcomes. Mutation is a genetic technique that helps prevent the premature conclusion to a suboptimal solution by diversifying an entire range of solutions. It helps new optimal solutions in solution sets that help the overall optimization process. Selection is an operator that chooses which solution from a current population fits enough to transition to the next phase. These drive genetic programming to find optimal solutions by ensuring the survival of only the fittest solutions in a set.

In soft computing, evolutionary computation helps applications of data mining (using large sets of data to find patterns), robotics, optimizing, and engineering methods. [3] [5]

Hybrid intelligence systems

Hybrid intelligence systems combine the strengths of soft computing components to create integrated computational models. Artificial techniques such as fuzzy logic, neural networks, and evolutionary computation combine to solve problems efficiently. These systems improve judgment, troubleshooting, and data analysis. Hybrid intelligence systems help overcome the limitations of individual AI approaches to improve performance, accuracy, and adaptability to address dynamic problems. It advances soft computing capabilities in data analysis, pattern recognition, and systems. [12]

Applications

Due to their dynamic versatility, soft computing models are precious tools that confront complex real-world problems. They are applicable in numerous industries and research fields:

Soft computing fuzzy logic and neural networks help with pattern recognition, image processing, and computer vision. Its versatility is vital in natural language processing as it helps decipher human emotions and language. They also aid in data mining and predictive analysis by obtaining priceless insights from enormous datasets. Soft computing helps optimize solutions from energy, financial forecasts, environmental and biological data modeling, and anything that deals with or requires models. [12] [13]

Within the medical field, soft computing is revolutionizing disease detection, creating plans to treat patients and models of healthcare. [10]

Challenges and limitations

Soft computing methods such as neural networks and fuzzy models are complicated and may need clarification. Sometimes, it takes effort to understand the logic behind neural network algorithms' decisions, making it challenging for a user to adopt them. In addition, it takes valuable, costly resources to feed models extensive data sets, and sometimes it is impossible to acquire the computational resources necessary. There are also significant hardware limitations which limits the computational power. [8]

Related Research Articles

Computer science is the study of the theoretical foundations of information and computation and their implementation and application in computer systems. One well known subject classification system for computer science is the ACM Computing Classification System devised by the Association for Computing Machinery.

<span class="mw-page-title-main">Evolutionary algorithm</span> Subset of evolutionary computation

Evolutionary algorithms (EA) reproduce essential elements of the biological evolution in a computer algorithm in order to solve “difficult” problems, at least approximately, for which no exact or satisfactory solution methods are known. They belong to the class of metaheuristics and are a subset of population based bio-inspired algorithms and evolutionary computation, which itself are part of the field of computational intelligence. The mechanisms of biological evolution that an EA mainly imitates are reproduction, mutation, recombination and selection. Candidate solutions to the optimization problem play the role of individuals in a population, and the fitness function determines the quality of the solutions (see also loss function). Evolution of the population then takes place after the repeated application of the above operators.

<span class="mw-page-title-main">Evolutionary computation</span> Trial and error problem solvers with a metaheuristic or stochastic optimization character

Evolutionary computation from computer science is a family of algorithms for global optimization inspired by biological evolution, and the subfield of artificial intelligence and soft computing studying these algorithms. In technical terms, they are a family of population-based trial and error problem solvers with a metaheuristic or stochastic optimization character.

Bio-inspired computing, short for biologically inspired computing, is a field of study which seeks to solve computer science problems using models of biology. It relates to connectionism, social behavior, and emergence. Within computer science, bio-inspired computing relates to artificial intelligence and machine learning. Bio-inspired computing is a major subset of natural computation.

Intelligent control is a class of control techniques that use various artificial intelligence computing approaches like neural networks, Bayesian probability, fuzzy logic, machine learning, reinforcement learning, evolutionary computation and genetic algorithms.

The expression computational intelligence (CI) usually refers to the ability of a computer to learn a specific task from data or experimental observation. Even though it is commonly considered a synonym of soft computing, there is still no commonly accepted definition of computational intelligence.

<span class="mw-page-title-main">Genetic fuzzy systems</span>

In computer science and operations research, Genetic fuzzy systems are fuzzy systems constructed by using genetic algorithms or genetic programming, which mimic the process of natural evolution, to identify its structure and parameter.

<span class="mw-page-title-main">Sankar Kumar Pal</span> Indian computer scientist

Sankar Kumar Pal is a computer scientist and the president of the Indian Statistical Institute (ISI), Kolkata. He is also a National Science Chair, Government of India. Pal is a computer scientist with an international reputation on pattern recognition, image processing, fuzzy neural network, rough fuzzy hybridization, soft computing, granular mining, and machine intelligence. He pioneered the development of fuzzy set theory, and neuro-fuzzy and rough-fuzzy computing for uncertainty modelling with demonstration in pattern recognition, image processing, machine learning, knowledge-based systems and data mining. This has made him widely recognized across the world and made India a leader in these disciplines in international scenario. He founded the Machine Intelligence Unit in 1993, and the Center for Soft Computing Research: A National Facility in 2004, both at the ISI. In the process he has created many renowned scientists.

<span class="mw-page-title-main">Neuro-fuzzy</span>

In the field of artificial intelligence, the designation neuro-fuzzy refers to combinations of artificial neural networks and fuzzy logic.

<span class="mw-page-title-main">Memetic algorithm</span> Algorithm for searching a problem space

In computer science and operations research, a memetic algorithm (MA) is an extension of an evolutionary algorithm (EA) that aims to accelerate the evolutionary search for the optimum. An EA is a metaheuristic that reproduces the basic principles of biological evolution as a computer algorithm in order to solve challenging optimization or planning tasks, at least approximately. An MA uses one or more suitable heuristics or local search techniques to improve the quality of solutions generated by the EA and to speed up the search. The effects on the reliability of finding the global optimum depend on both the use case and the design of the MA.

The following outline is provided as an overview of and topical guide to artificial intelligence:

Lateral computing is a lateral thinking approach to solving computing problems. Lateral thinking has been made popular by Edward de Bono. This thinking technique is applied to generate creative ideas and solve problems. Similarly, by applying lateral-computing techniques to a problem, it can become much easier to arrive at a computationally inexpensive, easy to implement, efficient, innovative or unconventional solution.

Natural computing, also called natural computation, is a terminology introduced to encompass three classes of methods: 1) those that take inspiration from nature for the development of novel problem-solving techniques; 2) those that are based on the use of computers to synthesize natural phenomena; and 3) those that employ natural materials to compute. The main fields of research that compose these three branches are artificial neural networks, evolutionary algorithms, swarm intelligence, artificial immune systems, fractal geometry, artificial life, DNA computing, and quantum computing, among others.

In computer science, an evolving intelligent system is a fuzzy logic system which improves the own performance by evolving rules. The technique is known from machine learning, in which external patterns are learned by an algorithm. Fuzzy logic based machine learning works with neuro-fuzzy systems.

An adaptive neuro-fuzzy inference system or adaptive network-based fuzzy inference system (ANFIS) is a kind of artificial neural network that is based on Takagi–Sugeno fuzzy inference system. The technique was developed in the early 1990s. Since it integrates both neural networks and fuzzy logic principles, it has potential to capture the benefits of both in a single framework.

This glossary of artificial intelligence is a list of definitions of terms and concepts relevant to the study of artificial intelligence (AI), its subdisciplines, and related fields. Related glossaries include Glossary of computer science, Glossary of robotics, and Glossary of machine vision.

The following outline is provided as an overview of, and topical guide to, machine learning:

Data-driven models are a class of computational models that primarily rely on historical data collected throughout a system's or process' lifetime to establish relationships between input, internal, and output variables. Commonly found in numerous articles and publications, data-driven models have evolved from earlier statistical models, overcoming limitations posed by strict assumptions about probability distributions. These models have gained prominence across various fields, particularly in the era of big data, artificial intelligence, and machine learning, where they offer valuable insights and predictions based on the available data.

References

  1. ↑ Zadeh, Lotfi A. (March 1994). "Fuzzy logic, neural networks, and soft computing". Communications of the ACM. 37 (3): 77–84. doi: 10.1145/175247.175255 . ISSN   0001-0782.
  2. 1 2 Ibrahim, Dogan. "An overview of soft computing." Procedia Computer Science 102 (2016): 34-38.
  3. 1 2 3 Kecman, Vojislav (2001). Learning and Soft Computing: Support Vector Machines, Neural Networks, and Fuzzy Logic Models. MIT Press. ISBN   978-0-262-11255-0.
  4. ↑ Chaturvedi, Devendra K. "Soft computing." Studies in Computational intelligence 103 (2008): 509-612.
  5. 1 2 3 Ram, Mangey; Davim, J. Paulo (2018-05-04). Advanced Mathematical Techniques in Engineering Sciences. CRC Press. ISBN   978-1-351-37189-6.
  6. ↑ "Fuzzy Logic | Introduction". GeeksforGeeks. 2018-04-10. Retrieved 2023-11-11.
  7. ↑ Trillas, Enric, and Luka Eciolaza. "Fuzzy logic." Springer International Publishing. DOI 10 (2015): 978-3.
  8. 1 2 Cheng, Yu; Wang, Duo; Zhou, Pan; Zhang, Tao (2018). "Model Compression and Acceleration for Deep Neural Networks: The Principles, Progress, and Challenges". IEEE Signal Processing Magazine. 35 (1): 126–136. Bibcode:2018ISPM...35a.126C. doi:10.1109/MSP.2017.2765695 . Retrieved 2023-11-11.
  9. ↑ "What are Neural Networks? | IBM". www.ibm.com. Retrieved 2023-11-11.
  10. 1 2 Abiodun, Oludare Isaac; Kiru, Muhammad Ubale; Jantan, Aman; Omolara, Abiodun Esther; Dada, Kemi Victoria; Umar, Abubakar Malah; Linus, Okafor Uchenwa; Arshad, Humaira; Kazaure, Abdullahi Aminu; Gana, Usman (2019). "Comprehensive Review of Artificial Neural Network Applications to Pattern Recognition". IEEE Access. 7: 158820–158846. Bibcode:2019IEEEA...7o8820A. doi: 10.1109/ACCESS.2019.2945545 . Retrieved 2023-11-11.
  11. ↑ "Genetic Algorithms". GeeksforGeeks. 2017-06-29. Retrieved 2023-11-11.
  12. 1 2 Medsker, Larry R. (2012-12-06). Hybrid Intelligent Systems. Springer Science & Business Media. ISBN   978-1-4615-2353-6.
  13. ↑ Dote, Y.; Ovaska, S.J. (2001). "Industrial applications of soft computing: a review". Proceedings of the IEEE. 89 (9): 1243–1265. doi:10.1109/5.949483 . Retrieved 2023-11-11.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.