Use of Artificial Intelligence in Decision Support Systems

Live Chat
Use of Artificial Intelligence in Decision Support Systems


This paper researches the way the artificial intelligence is built and how it is applied in the process of decision support systems’ functioning. A profound approach is used to depict the method of the artificial intelligence application in the form of algorithm. In this regard, the paper contains a comparison of how human brain and computer work. Hence, it was found that both the human brain and a computer have a similar structure of intelligence, but the speed of their processing is different – computers are much faster. In this regard, the application of the artificial intelligence is to be based on the interaction of a man and a computer to obtain the best result and the most appropriate solution. The paper also provides the examples of the artificial intelligence application in various fields of human expertise. As the experience shows, the artificial intelligence systems are very helpful and important elements of the decision support systems’ functioning.

Keywords: artificial intelligence, decision support system, system, algorithm, brain.

Use of Artificial Intelligence in Decision Support Systems

The concept of ‘artificial intelligence’ is used to refer to large areas of scientific and applied research. Such a term, assigned to this trend, is probably associated by most people with intelligent robots and intelligent computers, multiple images of which were created in the science-fiction works. Artificial intelligence development commenced in the 1950s as an independent branch of computer science. At the beginning, the scientists working in this field, set up a very ambitious goal which consisted in modeling of intellectual processes of the human brain. The first success in the intelligent automation of processes led to the formation of the leading and essential artificial intelligence paradigm, which also can be called a computer paradigm. The core of this idea is the assumption that the brain, from the point of view of information, is a large and complex biocomputer. That is why the intellectual processes which are necessary to work with the information on computer are based on symbolic representations of information and are carried out using algorithmic logic-combinatorial procedures. Soon, it became clear that for many intellectual tasks it is not necessary to copycat the brain processes. The high and ever-increasing speed of computing, surpassing the speed of brain signals by about a million times, allowed the use of complex sequential algorithms, which are not specific to the brain. Therefore, the paper proves the fact that artificial intelligence is one of the most suitable technologies to be used in the process of decision support system functioning.

Almost all the achievements of the last decades in artificial intelligence development have been made via the wide usage of intelligent technologies (such as expert systems, decision support technologies, and data mining) that are logic-combinatorial algorithms and usually have little to do with the processes of the brain. Accordingly, the computer paradigm has been transformed, as now it does not matter that there are significant differences between the brain and the computer processes; it is important that all intellectual problems can be solved by the computer (Hogenboom, Frasincar, Kaymak, de Jong, & Caron, 2016). However, over time, this thesis has been becoming less persuasive. Firstly, there are intellectual processes known as not amenable to computer simulation. They are creative processes, inspiration and intuition, the generation of various types of expert judgment, and other activities (Hogenboom et al., 2016). Secondly, many people resolve the intellectual tasks of the decision support in a mode which is faster than the computer’s, despite the fact that the speed of signal transmission in the brain is a million times lower than in the computer (Hogenboom et al., 2016). The most characteristic of two kinds of tasks occupy a special place among the cognitive processes (Hogenboom et al., 2016). A human being quickly recognizes not only individual objects, but also difficult situations. People also know how to develop a logically correct investigation, though they are not always able to build a valid proof.

The differences in the efficiency of computer and human intellectual problem-solving processes, decision support approaches and decision-making mechanisms can be simply expressed by the fact that what a person finds difficult, a computer can easily solve (Kasie, Bright, & Walker, 2016). All of this suggests that the study of cognitive structures and human mechanisms is important not only for the cognitive sciences, for which the brain processes are the direct object of study (such as, for example, psychology, neuroscience, and linguistics), but also for the artificial intelligence, which, using this research, may lead to the development of new intellectual technologies (Kasie et al., 2016). However, until now, not only virtually main artificial intelligence technologies represent the dominant idea of staying within the computer paradigm. At the same time, in the last decade, other cognitive sciences have emerged (Kasie et al., 2016). They comprise at least two notable concepts: cognitive semantics and the theory of the speaker are of direct interest to the artificial intelligence, but have not yet attracted enough attention of the corresponding experts.

Human knowledge is organized in the form of a conceptual system, which is a system of categories and relations between them. This system is structured radially: there are central (basic) and non-central categories and truths, which are important to be taken into account when discussing the role of artificial intelligence in decision support system. The central truths are characterized in terms of directly perceived notions, which are relevant to the pre-conceptual structure of the experience. They include basic categories of material field and schemes arising from the ordinary and professional experience (Saha, Aqlan, Lam, & Boldrin, 2016). Basic knowledge is acquired in the course of interaction with the environment through the perception and manipulation of material objects. The other types of knowledge are obtained either directly (emotional and social knowledge) or indirectly (the knowledge gained by others and transmitted through books, etc.). Based on this approach, the artificial intelligence is positively used in the system of decision support. This is due to its ability to form basic symbolic and image-schematic structure, correlating them with the pre-conceptual structures of our experience (Saha et al., 2016). In this regard, artificial intelligence also possesses the ability to construct the material structures in metaphorical projection areas using the structure of abstract domains due to structural correlations between them. Finally, artificial intelligence is able to form complex search using a structural scheme with shaped samples. Each of these aspects are of particular significance for the system of decision support.

The analysis of classical mathematical methods and tools of conducting research make decisions that have found very wide use in many application problems and subject areas. This leads to the conclusion that the traditional tools of the mid-20th century help make decisions and provide decision support when solving poorly structured and unstructured problems. It means that there is a need for the development and application of methods and models of artificial intelligence which are based on risk.

The problems which are associated with finding (construction) of the algorithm for solving a certain class of problems or making the decision and elaborating its support are called intellectual tasks (Saha et al., 2016). In respect to these classes of problems, the solving algorithms are already built, and then, they are attributed with the property of ‘intelligence.’ Indeed, after appropriate algorithm was built, the process of solving this class of problems is such that it can be accurately solved by a person and a computer (carefully programmed) or a controller, who does not have any idea about the essence (semantics) of the problem. For this, it is necessary to have only the person who solves the problem, and who would be able to perform the operations involved in the process. It is also significant to know that this process is meticulously and carefully guided by the proposed algorithm (Saha et al., 2016). Therefore, it is appropriate to remove such problems with intellectual class (for which there are standard algorithms and methods for the solution). Examples of such tasks may serve the following computing problems: solving a system of linear algebraic equations, numerical integration of differential equations and others. In contrast, for a variety of intellectual tasks such as pattern recognition, the game of chess, theorem proving there should be a formal allocation process of finding solution by breaking problem(s) into individual elementary operations (steps). This is often quite complex, even if their solution is not complicated. Hence, it is possible to rephrase the definition of artificial intelligence and its use in support of decisions systems in particular, as a universal algorithm, which is able to create other algorithms for solving certain classes of problems.

Intelligent decision-making system would ensure economists and managers have modern means of problem analysis and relevant information and would generate solutions. Their evaluation and selection of the best option should be based on the use of the above principles and the principle of irritation, implying the singled incremental synthesis and intelligent decision of the support systems (Saha et al., 2016). The majority of scholars believe that the integration of intelligent modeling tools of knowledge in decision-making, such as artificial neural networks, genetic algorithms, system on fuzzy logic and expert systems, will create a very powerful intelligent decision support system.

The structure of intelligent decision support system can highlight the following main subsystems. The first is the traditional intelligence systems modules, which include databases and knowledge; base models; mechanisms conclusions; system of accumulation and updating of knowledge; block explanations; organization of interactive user’s experience, and other elements (Kim & Eom, 2016). The second one comprises the following subsystems: analyzer’s problem situation; simulations of the problem situation; mathematical methods and models for analysis and forecasting.

One of the first attempts to create artificial intelligence systems was modeling experience in a relatively narrow professional credit sector. The experience includes cumulative knowledge and produced a system, which was based on the decision analysis methods of knowledge. Such systems are called expert, since their formation involved the most experienced specialists in the subject area, who are referred to as experts (Kim & Eom, 2016). Expert systems in which knowledge of many specialists and experts is accumulated, in fact, are capable of producing rational decisions (Saha et al., 2016). The problem is only in how to place expert knowledge in computer memory most adequately, to analyze this knowledge properly and to get new knowledge, which would be based on previous analysis.

The artificial intelligence systems are widely used to automatize the decision-making process in many areas of the economy. Economics as a science of economic management subjects requires a huge number of decisions for the effective implementation of numerous activities (Kim & Eom, 2016). The formalization of the process of decision-making is possible to recognize the principles of situations or objects of activities. Every situation can be characterized by the description that represents a specific set of attributes or properties for artificial intelligence systems’ input, and a full description of the further actions which are their output (Kim & Eom, 2016). Thus, the artificial intelligence systems provide recognition of the situation and a decision on further action by implementing expert systems.

The examples of tasks which can be completed with the artificial intelligence systems, may be the choice of the investment project, the investor, the carrier, the provider of choice, the choice of distribution channels of products, product quality evaluation, the formation of tariffs for transportation, depending on the characteristics of the product, and other tasks (Ngai, Lam, Poon, Shen, & Moon, 2016). Artificial intelligence systems can also be used in the operational work at the tactical level and in planning systems at the strategic level of optimal decision making.

The problem of recognition of objects (such as the identification of the economic situation at the macroeconomic level or loading units and vehicles at the micro level) requires a phased solution of many theoretical issues (Ngai et al., 2016). In the first phase, it is necessary to take into account that each object is described by the values of many properties or attributes, which form feature space (Ngai et al., 2016). Thus, it is impossible to determine information of the content of individual characters from the viewpoint of separability of classes of objects, and the required number of the most informative features. This model requires a change in the internal structure of the information within the artificial intelligence systems. It should be done in accordance with perspective and current tasks (Ngai et al., 2016). An indispensable condition for this is the existence of a strategic center, forming a strategy in a continuous mode. Hence, artificial intelligence systems are integrated into the planned economic information system on the basis of information-analytical center and can be used by participants of the economic processes in the region for their own purposes.

In conclusion, the paper regarded the role of the artificial intelligence system in decision support systems. To explain the difference between the tasks which require the use of artificial intelligence, one can use the term of algorithm. It is one of the cornerstones of mathematics and cybernetics. Algorithm is a clear instruction to perform a certain number of operations in a specific order to solve a particular problem with some class (set) of the problems. In mathematics and cybernetics, class of problems is a specific type, which is considered to be solved when their solution is installed (built) in the algorithm. Construction of the algorithm for solving problems of a certain type (class) is associated, in some cases, with subtle and complex reasoning, requiring ingenuity and skill. This kind of creative activity requires human intervention.

Go to our Order Page if you want to buy essay help on this or any other topic.