1 / 19

Decision-Making Techniques

Decision-Making Techniques. By: James Delk. What are Decision-Making Techniques(DMTs)?.

sharne
Télécharger la présentation

Decision-Making Techniques

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Decision-Making Techniques By: James Delk

  2. What are Decision-Making Techniques(DMTs)? • Decision making is the study of identifying and choosing alternatives based on the values and preferences of the decision maker. Making a decision implies that there are alternative choices to be considered, and in such a case we want not only to identify as many of these alternatives as possible but to choose the one that best fits with our goals, objectives, desires, values, and so on.. (Harris (1980)) Figure 1[1]

  3. What are DMTs cont. • “Decision making techniques fall into three major categories: random; intuition based; or analytical. ”[2]

  4. Requirements • Apples to oranges? • Assigning values Figure 2[3]

  5. DMT Setup • Define the problem • Determine requirements • Establish goals • Identify alternatives • Define criteria • Select the method to be used • Evaluate against criteria • Validate against problem[4]

  6. Why use DMTs • Physical representation of options • Double check variables • Future reference • Presentations • Standard of measurement • “if you don’t know where your going it doesn’t matter which way you go”(Lewis Carroll) Figure 3[5]

  7. DMT Example Pros vs. Cons (weighted) Figure 5[6]

  8. DMT Example Analytic Hierarchy Process Figure 6[2]

  9. DMT Example Simple Decision Tree • the number of possible decision trees is exponential in the number of attributes Figure 7[2]

  10. DMT Example Simple Influence Diagram Figure 8[2]

  11. DMT Example Linear Programming Figure 9[7]

  12. Computer Science? • AI • Databases • Programming • System design • Neural networks Figure 10[8]

  13. Computer Science cont. System architecture Figure 11[8]

  14. Game Theory • Game theory is "the study of mathematical models of conflict and cooperation between intelligent rational decision-makers.“[9] Zero-sum gain game theory A B A –1, 1 3, –3 B 0, 0 –2, 2 Figure 12[9]

  15. Multi Agent Systems • A multi-agent system (MAS) is a system composed of multiple interacting intelligent agents within an environment.[10] • Multi-agent systems are dependent upon a theoretical base that is provided through game theory. Multi-agent system Figure 13[10]

  16. MAS cont. • Allowing multiple units to break down computations to the easiest level for completion by many different agents Figure 14[11]

  17. Game theory sample code O((m+n)*iterations ''' Approximate the strategy oddments for 2 person zero-sum games of perfect information. Applies the iterative solution method described by J.D. Williams in his classic book, The CompleatStrategyst, ISBN 0-486-25101-2. See chapter 5, page 180 for details. ‘ '' from operator import add, neg defsolve(payoff_matrix, iterations=100): 'Return the oddments (mixed strategy ratios) for a given payoff matrix' transpose = zip(*payoff_matrix) numrows= len(payoff_matrix) numcols= len(transpose) row_cum_payoff= [0] * numrows col_cum_payoff= [0] * numcols colpos= range(numcols) rowpos= map(neg, xrange(numrows)) colcnt= [0] * numcolsrowcnt = [0] * numrows active = 0 for i in xrange(iterations): rowcnt[active] += 1 col_cum_payoff= map(add, payoff_matrix[active], col_cum_payoff) active = min(zip(col_cum_payoff, colpos))[1] colcnt[active] += 1 row_cum_payoff= map(add, transpose[active], row_cum_payoff) active = -max(zip(row_cum_payoff, rowpos))[1] value_of_game= (max(row_cum_payoff) + min(col_cum_payoff)) / 2.0 / iterations return rowcnt, colcnt, value_of_game Figure 15[12]

  18. Example AI process IBM’s Watson AI’s decision process http://www.youtube.com/watch?v=cU-AhmQ363I Figure 16[13]

  19. References • [1]hhttp://chamonixvue.wordpress.com/2012/02/06/decision-making/ • [2] http://www.decision-making-solutions.com/decision_making_techniques.html • [3]http://blogs.helsinki.fi/dervin/2012/03/13/cfp-appels-and-oranges/ • [4] Fülöp, J.: Introduction to Decision Making Methods. Working Paper 05-6, Laboratory of Operations Research and Decision Systems, Computer and Automation Institute, Hungarian Academy of Sciences, Budapest, November 2005 • [5]http://djcase.com/case-in-point/ginny-wallace/structured-decision-making-common-sense-made-explicit • [6] http://www.businessballs.com/problemsolving.htm • [7]http://www.cs.duke.edu/courses/spring08/cps296.2/ • [8]http://www.ece.ubc.ca/sites/default/files/Falessi%202011%20CSUR.pdf • [9] http://en.wikipedia.org/wiki/Game_theory • [10]http://en.wikipedia.org/wiki/Multi-agent_system • [11]http://fofoa.blogspot.com/2012/02/superorganism-open-forum.html • [12] http://code.activestate.com/recipes/496825-game-theory-payoff-matrix-solver/ • [13]http://en.wikipedia.org/wiki/File:DeepQA.svg

More Related