Computer support in human decision making

Key Words: Human decision making, multi criteria decision making methods, simplex method.

Abstract. Developments in computing offer experts in many fields specialized support for decision making under uncertainty. However, the impact of these technologies remains controversial. This paper present a mathematical method called "simplex method" which can be used in computer to both providing knowledge and information and also help humans to make final decisions with considerations of human factors. It concentrate the procedure of decision making, we get through the mechanism of decision making, both in human and by a system on the human side. We support this paper by first presenting some decision-Making Concepts in computer support decision making systems. At the end, we will present some examples for human of restrictions and relations formula that can be used in these kinds of model for better computer-aided human decision making.

Introduction

In recent years many researchers have focused on "complex problems" and computer-based decision systems that assist human decision makers. Complexity of a problem can be seen as function of the number and strengths of inter and intra relationships among internal and external components of the problem. Today, these complex problems can be found in many planning, operational, and management areas every fields that human have relation with like transportation, economics, engineering , military, medicine, and law.

Decisions support can improve the performance of human decision making by providing information and knowledge sources. It can also enhance the human decision maker to rationalize his decision by considering his favorite real-life factors into consideration in a complex and multi criteria ways. The difficulty of individual decisions and reliability of advice from the support tool depends on our way of looking at how the human makes decisions.

Human decision making is certainly not a very simple process. We do, therefore, not expect that it can necessarily be modeled by using very simple and computationally convenient operators. (Zimmermann 1978)But this is the price we have to pay in order to model features of human decisions more appropriate. Computers, including intelligent systems, assist human decision making in many ways (Hoffman Robert R 2010).

The difficulties which one faces so far when modeling human decisions and the lack of general acceptance of existing models is, due to the lack of transparency of information processing, which is performed in human brains before and when making decisions. An increased insight into these process can and will help to build more appropriate and more acceptable models of human decision making (Zimmermann H.J. 1980)

In this paper we discuss about computer aid in human decision making in another lever. After introducing computer aid providing knowledge and information source to help humans to make decision, we consider mathematical algorithms that can be used in computer software to help human make final decision easier in a multi criteria world.

To achieve a compromise between a computer support system and human decision, we will define some of the most important concepts in decision making.

Decision-Making Concepts

What's a decision?

The word "decide" has had much concentration in English, it can be somehow the definitions of the words arbitrate, conclude, convict, declare, define, disagree, intervene, or judge. "Decide" also have meaning to influence or determine. The modern English word "decision" comes from the Sanskrit "khidati" meaning to tear. We see here a notion that the decision is a point, and a final point or action separating one thing from another (historically, a human head or limb from its body).

All dictionary definitions "decision" as bringing a series of events, including a mental sequence, to a final and conclusion point. In this paper we consider "decision" as an act of decision maker to do or not to do a certain task or act.

Decision Making

Decision making is an outcome of mental processes (cognitive process). It can lead to the selection of a certain action among several available alternatives. Every decision making process has a final result (choice). The result can be do or not to do an action or an alternative of choice.

Decision making terms has been the subject of active research from several perspectives with consideration of human performance. From a psychological perspective, it is necessary to understand individual decisions to determine set of needs, preferences and values or thoughts of every person.

So decision making can be defined as an action that has the intention of maximizing the interests and values of particular person.

A decision is a mental event that occurs at a specific time, a psychological moment of choice that leads immediately or directly to do or not to do an action (for example, going to Metallica's concert, or not call the one you have seen yesterday at bar).

Most views of decision making are in fields of psychology and of judgment.

Decision making determine the notion of decision in the stage-theoretic framework in it's final stage. In the final-point notion we can say decisions are made, but even in apparently simple cases; the deciding process is much more than this. (Yates 2006)

Decision Theory

Identifying the values, uncertainties and other issues relevant in a given decision, its rationality, and the resulting optimal decision has always been a topic in philosophy, mathematics and statistics.

Most of decision theory is normative or prescriptive. They try optimizing a decision for a take, Decision theories assume that the decision maker is fully informed, able to compute and is rational. But rationality is not something unique. Each person has his own rationality that can be irrational in view of similar peer decision maker. Software tools that have been developed in this way are called decision support systems. They try to look systematic and comprehensive in human decision making process and results.

Since people usually do not behave in ways consistent with norms, often their own. Positive or descriptive discipline there is a related area of study to discuss in this subject. Scientists in this area try to describe what people will actually do. Since the normative, optimal decision often creates hypotheses for testing against actual behavior, the two fields are closely linked.

So it is possible to look at the assumptions of perfect information or rationality in various ways. It produces different prescriptions and predictions about behavior and will allow further tests to be compared with decision-making that occurs in practice. (Wikipedia 2010)

Cognition

Cognition is considered "the process of thought" to knowing in scientific fields. Meaning of cognition varies in different fields, for example in psychology and cognitive science; it usually is an information processing view of an individual's psychological functions. Other meaning of cognition connects it to the development of concepts, individual minds, groups and organizations.

The psychological output of cognition is perception and learning and reasoning. Cognition is related to the social and human organization functioning and constrains specially in systemic perspective. Managerial decision making processes can many errors in politics, economy and industry. The main reason is different reciprocally dependent socio-cognitive factors. Recently, Cognition has become the field of interest of emergent socio-cognitive engineering. (Piaget's 2010)

In the cognitive perspective, the decision making process is considered continuous process that is integrated and have interaction with the environment. In the normative perspective, the analysis of individual decisions is related to logic of decision making and rationality. (Wright G 1992)

Neuroscience perspective

Two main brain regions which are involved in decision making processes are:

  1. The anterior cingulated cortex (ACC) and
  2. Orbit frontal cortex (and the overlapping ventromedial prefrontal cortex).

A recent neuroimaging study (Kennerley W 2006) found distinctive patterns of neural activation in these regions .that study shows that whether decisions were made on the basis of personal volition or following directions from someone else. Other studies have shown that Patients with damage to the ventromedial prefrontal cortex have difficulty making advantageous decisions (AR 1994)

A recent study( Kiani 2009) involving Rhesus monkeys found that neurons in the parietal cortex is representing the formation of a decision and signal the degree of certainty which is associated with the decision.

Other researchers found that lesions to the ACC in the macaque resulted in impaired decision. Studies show that ACC may be involved in evaluating past reinforcement information and guiding future action. Emotion of each person influence the decision making process. Decision making has always uncertainty about whether one's decision will lead to benefit or loss. The somatic-marker hypothesis shows how decisions are made with uncertain outcome.

This theory describes that emotions influence such decisions, the mind consider emotions, deliberate of future consequences and mark different options for behavior as advantageous or disadvantageous. This process has interplay between neural systems that evoke emotional/bodily states and neural systems that map these states (Naqvi 2006).

Human - Computer Interaction

Human - Computer interaction (HCI) is the study that is focused on interaction between people and computers. It is somehow intersection of computer science, behavioral sciences and design. Interaction between users and computers happens at the user interface (interface). Both software and hardware must be involved. For example, all objects which are displayed by software on a computer's monitor, input received from users via hardware such as keyboards.

Human-computer interaction is define as "a discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them" by Association for Computing Machinery (Bohanec, 2010).

Human-computer interaction considers a human and a machine in conjunction, knowledge on both the machine and the human side support it. From the machine perspective, techniques in operating systems, programming languages, and development environments are relevant. From the human perspective, social sciences, industrial design disciplines, psychology, communication theory , Engineering and design methods linguistics and human factors are relevant.

HCI is also defined as man-machine interaction (MMI) or computer-human interaction (CHI). Poorly designed human-machine interfaces can cause many unexpected problems so Attention to human-machine interaction is so important. For example, investigations in Three Mile Island accident have shown that the design of the human-machine interface was responsible for the disaster (Webster's Dictionary 1979)

Very similar to that, in accidents in aviation, even though the new designs were proposed to be superior in comparison to basic human-machine interaction, pilots had already established the "standard" layout. Thus the conceptually good idea actually had bad results. Thus, human-machine interaction is considered a complex field of practical study.

Intelligence

Intelligence s is a property of the mind that has the abilities and capacities for learning, planning and problem solving, abstract thought, reasoning and the use of language.

Theories of intelligence are:

  1. The "single intelligence" which is based on the uni linear construct of "general intelligence"
  2. The construct of multiple intelligences. Francis Galton proposed a theory of general intelligence. He stated that intelligence is a biologically-based mental function that can be studied by measuring a person's reaction times to cognitive tasks. Galton measured the head sizes of scientists and laymen. He has concluded that head-size is unrelated to a person's intelligence.

Alfred Binet , has considered that intelligence is a median average of dissimilar abilities, not a unitary entity with specific, identifiable properties. The theorists of general intelligence and multiple intelligence use the Binet test in measuring intelligence. Evidence shows that genetic variation has a significant impact on human intelligence. They also demonstrated superior ability to learn and have the memory in various behavioral tasks (Pomerol 1997).

Collaborative intelligence

Collaborative intelligence is a measure of the collaborative ability of a group or entity. Today, knowledge of collaborative efforts is specially World Wide Web.

Collaborative intelligence is interaction between groups that collaborate within a virtual sphere. Group members can interact in real time even though they are not in the same physical space at that time. (thinkquest.org 2010)

Critical success factors for a high collaborative intelligence quotient are: Group moderation and facilitation ,Adherence to a small set of fundamental rules relate to member interaction ,No limits to thinking; or the promotion of creative thinking ,Strong group membership feedback, Quality control, Ideas need to be nurtured, but the solutions should be upheld after a critical peer review, The construction of a deeply documented group memory or knowledge base (Pomerol 1997)

Research into the areas of learning, language, and perception of human helped us to build intelligent machines. The most challenge that experts faces, is building a systems that simulate the behavior of the human brain. It should be made up of billions of neurons.

One of the ways to evaluate the intelligence of a machine is Alan Turing's test . He believes that a computer can be called intelligent if it could deceive a human into believing that it (the machine) was human. (Marsh 2003).we should consider that when a task is programmed, the decision no longer exists since the actions are determined according to each possible situation that may occur (Pomerol 1997)

Human vs. Machine Decision Making

Beside the human beings, there are many artificial systems that also make decisions: switching circuits, computer programs, autonomous expert systems and software agents and many else. Therefore, we differentiate between machine and human decision making.

The two disciplines that correspond to this distinction are Decision Systems and Decision Sciences. Decision Systems deals with computer-based programs and technologies intended to make decisions, monitor and control processes ( Power 2002). Decision Sciences is a broad discipline about human decision making.

Multi criteria decision making

Multiple Criteria Decision Making was introduced as an important field of study in the early 1970. There are two kinds of view to see Human-decision making:

  1. The outcome-oriented approach
  2. This kind of view is based on the view that if a person predicts the outcome of decision process, he understands the decision process. Result of this process and precision of the outcome is the main subject in this kind of view.

    The practical application of this prescriptive approach (how people actually make decisions) is called decision analysis, and aimed at finding tools, methodologies and software to help people make better decisions(Wikipedia 2010).

  3. The process-oriented approach
  4. This kind of view is based on if we can understand the decision process, then we can correctly predict the outcome. It is only when we face many attributes, criteria, functions, alternatives, factors and objectives, we can consider decision making and the theories which is related to it. In this paper we develop this kind of view with presenting symbolic equivalent of human decision making process by a mathematical model with can be used in a computer to simulate and assist human decision maker. It can improve the likeliness of the decision for the decision maker in a complex situation.(Zeleny M 1982).

Presented model

We have considered human decisions into two types:

Making a choice between different choices (Do or not do to) or in another words prioritizing task to do, for example choosing between going to different locations for trips. Every location is a choice for a person, he should say yes to one location and no to the other ones. To quantities the problem, we will consider for each choice a multiplier. In this kind of work the multipliers is other 0 or 1. The 0 means that the location is not selected by the person and 1 means the person has selected this location to go to trip. We will formulize this human decision like this:

Choice of the person= {0 or 1}*(alternative1)+ {0 or 1}*(alternative 2)+............

We can consider prioritizing of chooses in human minds also in here, we will assume that after prioritizing the choices in human mind, person will choose the first choice with the most privilege. For more elaborating, we consider the factors of decision making for every person into account.

Factors (variables) of decision making

Each human makes his decisions by considering some factors (variables) for it, the multi criteria decisions making technique look to process of decision making this way. We can consider everything that may influence or decision as a factor like: intuition, past memories, religion thoughts, economic issues and...

We can consider even "irrationally" of the person as a factor in our algorithms. This can solve the problem when we consider unmoral people. Influence of this factor depends on the multiplier of it for every person that we will describe in next part. These factor makes a final goal formula in every human brains.

Goals and Irrational persons

We will consider that every human being will consider a goal in making a decision. Choosing to do or not to do an action, making comparison or prioritize thing, for all this examples, the decision maker should have a goal. In this paper we consider a goal for each decision maker.

All the things that we do have an intention (we have consider it as an factor), even if we say that there is a person who does thing without any intention or goal, we can consider the factor of "irrationality" for this person that can change the whole balance of formula for decision making but the result will something that we will call irrational, but for that special person the result is completely rational in his view and he has reached his goal. Irrational person will try to optimize his goal that will be "more irrationality", so there will be no conflict with or model.

Objectives must first be established, then classified and placed in order of importance for the person that will make the decisions, and then the choices will be evaluated against each other's (Kepner 1965)

For using this kind of approach the best way that suits our explanation and assumption is Operations Research methods.

Operation research

Operations research is a method of mathematics science that uses quantitative analytical methods to arrive at optimal for decision-making problems.(Ackoff R.L 1968). This method was presented in British army during the WW II. Operation research methods will determine the maximum (e.g. profit, preference performance) or minimum (e.g. loss, risk) of some real-world objective. (Nacht M 2004)

One can consider a "human decision making system" as the inter-relationship among the components or functional units. Operations research seeks to find the best solution available to achieve the system's objectives considering the constraints under which the system operates. We have presented the Weight factor and mathematical restriction to show these meanings.

At first , we present a simple view to look at human decision making in operation research and then we will elaborate it more and use "Simplex method" for it.

Simple view

In simple view, the person compares alternatives. Each of these alternatives considers the factor that have an influence on the person decision. To formulate this, person compares:

{(alternative1)*(factor 1) + (alternative1)*(factor 2) +...+ (alternative1)*(factor N)}

With

{(alternative2)*(factor 1) + (alternative2)*(factor 2) +...+ (alternative2)*(factor N)}

(Alternative M)*(factor 1) + (alternative M)*(factor 2) +...+ (alternative M)*(factor N)}

And the decision maker will choose the alternative that has gain the most credits among the others.

More detailed view

Looking into more details should enable us to take more restriction and relation into consideration. This can improve our model to be more consistent to reality and simulation of every human behavior in decision making in reality.

For this kind of view we propose to use "simplex method" to better formulate the process and nature of decision making by every human.

The Simplex method

The Simplex method or Linear programming is a mathematical optimization theory Consider a linear programming problem, and tries to optimize (maximize) a goal with considering the constraints. Simple view of formulating a problem with simplex method is like this:

Maximize CX ( personal goal)

Subject to

Are the variables of the problem

Are the profits or costs or preferences multiplier for each variable in the goal

Are the linear constraints

We consider the goal of every person in his mind like this. We consider that in decision maker mind there are N factors and there are M alternative for him to choose:

Decision maker Goal =

Max

{(alternative1)*( weight factor 1)+ (alternative1)*(weight factor 2)+....

+ { (alternative2)*( weight factor 1)+ { (alternative2)*( weight factor 2)+..........

+ { alternative M)*( weight factor N)}

The simplex method will calculate every possibility of this formula and will choose the best amount for every alternative (that can be 0 or 1) which make the goal maximum.

Wight factors and restrictions:

Every human considers many factors for his selection to do or not to a task. In his mind, he will prioritize these factors. This can be conscious or unconscious. We will assume that Human mind can compare every factor and considers one factor bigger or smaller or equal to others.

In this paper we consider the relation between alternatives, selection factors, or restriction in every human decision maker in restriction part. Then simplex method will solve this mathematical problem with its method and will give the results which will be the amount of 1 for one alternative (the chosen alternative)and 0 for the rest. Factor can be or + and decision maker can define the range of this weight factor in his mind.

Here we describe some methods for describing the relations and restrictions in quantities ways:

  • Alternatives= {0 or 1}

This means that we the decision maker can do or not to do a certain task or choosing certain alternatives.

The aggregation of subjective categories in the framework of human decisions or evaluations almost always shows some degree of compensation. This indicates that human beings partially are using a procedures which do not correspond to the verbal and logical connectives "and" and "or" or (0 or 1). Several well-known operators are tested. However, they do not predict human's decision situation very well. Therefore a new class of operators is suggested which varies with respect to a parameter of compensation. The fuzzy system simply generalized the choice between 0 and 1 ( be or not to be) to choice from a range of [0,1]. (Zadeh, 1976) A general evaluation of the potential of this approach is not possible before empirical results are available.

  • Alternative1+ alternative2+.....+alternative N = 1

To impose this rule that decision maker should select only one alternative among all of alternatives.

  • Weight Factor X < or > or = Weight Factor Y

Comparing the importance of each Wight Factor to other ones can be show by this.

  • Factor X +factor Y = 1

If two alternatives have contradiction with together for the decision maker and he prefers to take only one of them into consideration, it can be formulated like this.

  • Irrationality, intuition, spiritual beliefs

Factor X = 8

Presented model has ability to take into consideration Irrationality, intuition, spiritual beliefs of every person as a factor for making decision. As we define intuition as an internal signal that can influence whole normal rationality and logic of human, we consider 8 for amount of these kinds of factors. So if the person considers his intuition, spiritual or religious beliefs, the weight factor is big enough to change the whole formula and then every alternative will have the rank of 1 in simplex method. It means that with considering this factor each alternative can be selected. So the decision will depends on what that religion or spiritual meaning say to the person.

Actually in this case, there will be different alternative at first because the decision maker has already made a preselecting for his choice and the factor is strong enough that not all other factor can have an influence on the final decision.

These are just some examples of how we can consider relations and restrictions. It can be more elaborate to give closer accuracy to decisions that will be made in reality.

Conclusions

21st century information technology will improve situational awareness and understanding far beyond the classic decision making, but these impacts are useful only if human can use that awareness and understanding to make decisions better and faster than the ever before. The aim is to reach "decision superiority" not just information superiority. In the fast-paced future, humans will rely more and more on information technology to deliver knowledge and to assist them in using that knowledge the decisions will have to be reached by a mix of human and machine reasoning, including a theater-wide perspective as well as a local perspective.

In this paper we presented a mathematical method called "simplex method" which can be used in computer to both providing knowledge and information and also help humans to make final decisions with considerations of human factors.

With help of this kind of view and method, human decision maker can be use computer software simulator to help taking the decisions easier in a multi criteria world. With more elaboration in this model, we can improve the likeliness of the decision for the decision maker in a complex situation.

However ,what was said about this proposed model, is also true for other human decision making models, Goal programming and Multi criteria decision making might become more convincing if the assumption of these models would have been tested in practical situation and to be considered carefully to their appropriateness, this can be a broad field for further research on this area.

After all, we should take into consideration that decision making is dynamic process. A complex system with many detours, continuous updated information, and especially many uncertainties. So there is a long road to understand all of these thing and formulate them to make a near-reality situation for decision making in a computer. We should consider that man is a reluctant decision maker, not a computer or machine.

Please be aware that the free essay that you were just reading was not written by us. This essay, and all of the others available to view on the website, were provided to us by students in exchange for services that we offer. This relationship helps our students to get an even better deal while also contributing to the biggest free essay resource in the UK!