Introduction To Decision Modeling
Introduction to Decision ModelingBrice MayagUniversity Paris DauphineLAMSADEFRANCEChapter 0Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 01 / 36
Outline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 02 / 36
ModelsOutline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 03 / 36
ModelsCurrent definitions of a modela standard or example for imitation or comparison.a representation, generally in miniature, to show the construction orappearance of something.an image in clay, wax, or the like, to be reproduced in more durable material.a person or thing that serves as a subject for an artist, sculptor, writer, etc.a person whose profession is posing for artists or photographers.a person employed to wear clothing or pose with a product for purposes ofdisplay and advertising.a style or design of a particular product: His car is last year’s model.a pattern or mode of structure or formation.a typical form or style.a simplified representation of a system or phenomenon, as in the sciences oreconomics, with any hypotheses required to describe the system or explainthe phenomenon, often mathematically.Zoology: an animal that is mimicked in form or color by another.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 04 / 36
ModelsWhat is a model? Representation of realityMore Precisely: A model refers to some form of symbolic representation ofour assumptions about realityWhy do we use models?Enhance our understanding of the world to improve our decision makingElaborate a scientific method to solve a problemDuplicable (repeatable)Reduce biasBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 05 / 36
ModelsTypes of models1Deterministic modelsoutcomes are precisely determined through known relationships among statesand eventsin such models, a given input will always produce the same outputEx: Resources to make a PC are the same every timeDomains: Multi-Attribute Decision Making; Linear programing; . . .2Probabilistic (stochastic) modelsNot all data is known with certaintyEx: College acceptance, being above average increases likelihood ofacceptance but does not make it certainDomains: Queuing; Simulation; . . .Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 06 / 36
ModelsModels are fed by data1Qualitative datameasured by qualityExpert opinionsEx: class atmosphere, . . .2Quantitative dataEasily measured by numbersEx: Numbers of tv programs a day; number of applications in a phone; . . .Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 07 / 36
ModelsModels are used every dayA Scenario: Driving to schoolAt what time do you need to leave home to be at school on time?Distance Rate TimeTime Distance/RateBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 08 / 36
ModelsFormal models vs Informal modelsA formal model is a precise statement of components to be used and therelationships among them.Formal models are usually stated via mathematics, often equations.Formal models can be precisely communicated because they are well-defined.Formal models give replicable results. This is the simple meaning of“mathematical proof”.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 09 / 36
ModelsFormal models vs Informal modelsFormal models are not reality: you must choose the model.Formal models may not correspond to reality: the prediction will turn out tobe false.An informal model is one in which the symbols are mental, verbal, orpictorial: e.g. we toss a coin, we ask an oracle, we visit an astrologer, weconsult an expert, we thinkInformal models simply have some lack of precision. Some relationships maynot be stated as equationsBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 010 / 36
Decision theory and Decision analysisOutline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 011 / 36
Decision theory and Decision analysisA definition of DecisionThe act or process of deciding; determination, as of a question or doubt, bymaking a judgment:Ex: They must make a decision between these two contestants.The act of or need for making up one’s mind:This is a difficult decision.Something that is decided; resolution:Ex: She made a poor decision when she dropped out of school.A judgment, as one formally pronounced by a court:Ex: It is the decision of this court that the appeal is granted.The quality of being decided; firmness:Ex: He spoke with decision and calm authority.Source: http://www.dictionary.com/browse/decisionBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 012 / 36
Decision theory and Decision analysisProvisional definition of Decision [RONALD HOWARD]“Decision-making is what you do when you do not know what to do”Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 013 / 36
Decision theory and Decision analysisA definition of Decision in our contextA choice that you make about something after thinking about severalpossibilitiesEx: We need to take a lot of factors into account in our decision-making.Ex: She has had to make some very difficult decisions.Ex: The company will reach/come to/make a decision ctionary/english/decisionBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 014 / 36
Decision theory and Decision analysisDecision in many domainsPhilosophy, Economics, Mathematics, Operational Research, Psychology,Computer sciences, Political sciences, Biology? Theology?Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 015 / 36
Decision theory and Decision analysisWhat Decision Analysis is not !A general method for taking “good decisions”Example 1: Choice of new jobExample 2: medical decisionEtc.What is a “good decision”?Good for whom, according to what criteria, at which moment in time?Good decision processes vs. good decisions?A description on how “wise people” decideExpert systemsDoctors / Politicians: Nuclear Industry vs Road safety; Prevention vs First AidHow do you recognize “wise people”?Luck vs. WisdomBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 016 / 36
Decision theory and Decision analysisDecision AnalysisDefinition (B. Roy): “consists in trying to provide answers to questions raisedby actors involved in a decision process using a model”Decision process: strategy of intervention: aid, communication, justification,etc.Many different ways to provide decision-aidDifficulty to compare methodsWhat is a “good” Decision Analysis model ?Different models may lead to different recommendationsBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 017 / 36
Decision theory and Decision analysisDecision AnalysisDefinition (B. Roy): “consists in trying to provide answers to questions raisedby actors involved in a decision process using a model”Answers: “Optimal solution” or “Good decision” is absentModels: formalized or notBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 018 / 36
Decision theory and Decision analysisDecision MakingDecision Making 6 “Solving”a well-defined problemIntervention in a decision process:imagine duct changeImportance of “final choice” ?Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 019 / 36
Decision theory and Decision analysisFormal decision modelsA set of explicit and well-defined rules to collect, assess and processinformation in order to be able to make recommendations in decision and/orevaluation processesA perfect or not even a best formal decision model do not exist.It is important to describe the decision model used (transparency?).Actually, defining a “perfect model” would be a difficult, if not impossible,task.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 020 / 36
Decision theory and Decision analysisExamples of modelsAstrologythe astrologer “provide answers to questions raised by his/her client using amodel”GraphologyPsycho-analysisDecision analysis makes use of explicit and formalized modelsBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 021 / 36
Decision theory and Decision analysisFormalized modelsDrawbacks: complex, opaqueAdvantages:Provide a clear language: communication toolCapture the essence of a situation: structuration toolAnswers “what-if” questions (sensitivity, robustness): Exploration toolExample: choosing a bottle of wineBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 022 / 36
Decision theory and Decision analysisPossible objectionsI do not need such tools because I know how to decideLet’s organize a high-level meeting to discuss itIntuition is often enoughBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 023 / 36
Main steps of developing a decision modelOutline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 024 / 36
Main steps of developing a decision modelDeveloping a decision model (Step 1)Formulation: Translate the problem scenario into a mathematical modelDefine the problemDevelop a decision modelVariables: Measurable quantity that can be variableParameters: measurable quantity inherent to problemBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 025 / 36
Main steps of developing a decision modelDeveloping a decision model (Step 2)Solution: Mathematical expressions from formulation are solvedDevelop a Solution: Manipulate model to arrive at the best solution. Ex:Time Distance/RateTest Solution: Does the solution make sense?Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 026 / 36
Main steps of developing a decision modelDeveloping a decision model (Step 3)Interpretation: Implication of resultsConduct sensitivity analysis:what happens if parameters vary?Testing outcomes under a variety of statesImplement results: Enact solution & monitor it performs as expectedBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 027 / 36
Main steps of developing a decision modelPossible problemsPossible problemsDefining the problem: Conflicting viewpoints, impact on other stakeholdersModel development: Fitting problem scenario to textbook model,understanding of othersAcquering data: Existence, validityDeveloping a solution: Limitations of one answerImplementation: Management and user supportBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 028 / 36
Decision’s algorithm & TransparencyOutline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 029 / 36
Decision’s algorithm & TransparencyDecisions made by algorithms can be opaque because of technical and socialreasons, in addition to being made purposely opaque to protect intellectualproperty.For example, the algorithms may be too complex to explain or efforts toexplain the algorithms might require the use of data that violates a country’sprivacy regulations.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 030 / 36
Decision’s algorithm & TransparencyAn algorithm : DefinitionAn algorithm is a sequence of instructions, typically to solve a class of problems orperform a computation.An algorithm has to beFinite: The algorithm must eventually solve the problem;Well-defined: The series of steps must be precise and present steps that areunderstandable;Effective: An algorithm must solve all cases of the problem for whichsomeone defined it.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 031 / 36
Decision’s algorithm & TransparencyAn algorithm : DefinitionObjectives sometimes contradictoryAn algorithm has to be simple to understand, to implement;The computation time of an algorithm should be reasonable.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 032 / 36
Decision’s algorithm & TransparencyTransparencyAlgorithmic TransparencyAlgorithmic transparency is the principle that the factors that influence thedecisions made by algorithms should be visible, or transparent, to the peoplewho use, regulate, and are affected by systems that employ those algorithms.Algorithmic transparency is openness about the purpose, structure andunderlying actions of the algorithms used to search for, process and deliverinformation.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 033 / 36
Decision’s algorithm & TransparencyTransparencyTwo important propertiesExplanation: Systems and institutions that use algorithmic decision-makingare encouraged to produce explanations regarding both the proceduresfollowed by the algorithm and the specific decisions that are made. This isparticularly important in public policy contextsAccountability: Institutions should be held responsible for decisions made bythe algorithms that they use, even if it is not feasible to explain in detail howthe algorithms produce their results.Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 034 / 36
Our programOutline1Models2Decision theory and Decision analysis3Main steps of developing a decision model4Decision’s algorithm & Transparency5Our programBrice Mayag (LAMSADE)Introduction to Decision ModelingChapter 035 / 36
Our programOur ProgramChapter 0: Introduction to Decision ModelingChapter 1: Preferences as binary relationsDefinitionNumerical representationExercises Practical works: Binary relations properties; Preferences forholidaysChapter 2: Measurement and meaningfulnessChapter 3: Group decision-makingSocial choice theory modelsExercises Practical works: Voting methods, Collaborative filtering;Chapter 4: Multi-Criteria Decision Aid modelsMulti-Attribute values approachesOutranking approachesExercises Tutorial: Decision Deck platformProject: Elaboration of an understandable evaluation model based on anMCDA approach: a raking of hospitals or hotels or . . .Final Evaluation: Practical works (20 %) Test (30%) Project (50%)Brice Mayag (LAMSADE)Introduction to Decision ModelingChapter 036 / 36
Decision theory and Decision analysis Decision Analysis De nition (B. Roy):\consists in trying to provide answers to questions raised by actors involved in a decision process using a model" Answers:\Optimal solution" or \Good decision" is absent Models:formalized or not Brice Mayag (LAMSADE) Introduction to Decision Modeling Chapter 0 18 / 36
14 D Unit 5.1 Geometric Relationships - Forms and Shapes 15 C Unit 6.4 Modeling - Mathematical 16 B Unit 6.5 Modeling - Computer 17 A Unit 6.1 Modeling - Conceptual 18 D Unit 6.5 Modeling - Computer 19 C Unit 6.5 Modeling - Computer 20 B Unit 6.1 Modeling - Conceptual 21 D Unit 6.3 Modeling - Physical 22 A Unit 6.5 Modeling - Computer
work/products (Beading, Candles, Carving, Food Products, Soap, Weaving, etc.) ⃝I understand that if my work contains Indigenous visual representation that it is a reflection of the Indigenous culture of my native region. ⃝To the best of my knowledge, my work/products fall within Craft Council standards and expectations with respect to
Structural equation modeling Item response theory analysis Growth modeling Latent class analysis Latent transition analysis (Hidden Markov modeling) Growth mixture modeling Survival analysis Missing data modeling Multilevel analysis Complex survey data analysis Bayesian analysis Causal inference Bengt Muthen & Linda Muth en Mplus Modeling 9 .
Oracle Policy Modeling User's Guide (Brazilian Portuguese) Oracle Policy Modeling User's Guide (French) Oracle Policy Modeling User's Guide (Italian) Oracle Policy Modeling User's Guide (Simplified Chinese) Oracle Policy Modeling User's Guide (Spanish) Structure Path Purpose Program Files\Oracle\Policy Modeling This is the default install folder.
1. What is decision theory?.5 1.1 The decision disciplines 5 1.2 Decision processes 7 1.3 Decision matrices 11 1.4 Classification of decision theories 13 1.4.1 Normative and descriptive theories 14 1.4.2 Individual and collective decision-making 15 1.4.3 Degrees of knowledge 16 2.
Oct 18, 2014 · A decision problem is characterized by decision alternatives, states of nature, and resulting payoffs. The decision alternatives are the different possible strategies the decision maker can employ. The states of nature refer to future events, not under the control of the decision maker, which
tables syntax and layout are defined by the DMN standard while Drools native decision tables are defined by the Drools project. Red Hat Decision Manager supports both formats of decision tables, but they are not interchangeable. For more information about Drools decision tables, see Designing a decision service using uploaded decision tables. 1 .
American Revolution This question is based on the accompanying document (1-6). The question is designed to test your ability to work with historical documents. Some of the documents have been edited for the purposes of the question. As you analyze the documents, take into account the source of each document and any point of view that may be presented in the document. HISTORICAL CONTEXT: passed .
