Process Model Realism Measuring Implicit Realism 8/09/2014dr. Benoît Depaire.

Slides:



Advertisements
Similar presentations
DATA & STATISTICS 101 Presented by Stu Nagourney NJDEP, OQA.
Advertisements

II. Potential Errors In Epidemiologic Studies Random Error Dr. Sherine Shawky.
MATHCOUNTS TOOLBOX Facts, Formulas and Tricks
CHAPTER 14: Confidence Intervals: The Basics
1 Counting Techniques: Possibility Trees, Multiplication Rule, Permutations.
1 VLDB 2006, Seoul Mapping a Moving Landscape by Mining Mountains of Logs Automated Generation of a Dependency Model for HUG’s Clinical System Mirko Steinle,
Lecture (11,12) Parameter Estimation of PDF and Fitting a Distribution Function.
Community and gradient analysis: Matrix approaches in macroecology The world comes in fragments.
FAO assessment of global undernourishment. Current practice and possible improvements Carlo Cafiero, ESS Rome, September CFS Round Table on.
Using Divide and Conquer for Sorting
Introduction to Probability
Types of Group Designs _________-subject design. The experiment compares _____ group across different levels of the IV. e.g., behavior is studied in 1.
Power Laws Otherwise known as any semi- straight line on a log-log plot.
Lecture 07 Prof. Dr. M. Junaid Mughal
Introduction to Communication Research
PSY 307 – Statistics for the Behavioral Sciences
Part I: Classification and Bayesian Learning
Slide 9-1 © The McGraw-Hill Companies, Inc., 2006 Audit Sampling.
Repeated Measures Designs
System/Software Testing
©2010 Prentice Hall Business Publishing, Auditing 13/e, Arens//Elder/Beasley Audit Sampling for Tests of Details of Balances Chapter 17.
©2012 Pearson Education, Auditing 14/e, Arens/Elder/Beasley Audit Sampling for Tests of Details of Balances Chapter 17.
C ONFORMANCE C HECKING OF P ROCESSES B ASED ON M ONITORING R EAL B EHAVIOR Jason Ree 4/18/11 UNIST School of Technology Management.
Basic Business Statistics, 11e © 2009 Prentice-Hall, Inc. Chap 8-1 Confidence Interval Estimation.
1 Chapter 6. Section 6-1 and 6-2. Triola, Elementary Statistics, Eighth Edition. Copyright Addison Wesley Longman M ARIO F. T RIOLA E IGHTH E DITION.
Population All members of a set which have a given characteristic. Population Data Data associated with a certain population. Population Parameter A measure.
Statistical Thermodynamics Chapter Introduction The object: to present a particle theory which can interpret the equilibrium thermal properties.
4 Hypothesis & Testing. CHAPTER OUTLINE 4-1 STATISTICAL INFERENCE 4-2 POINT ESTIMATION 4-3 HYPOTHESIS TESTING Statistical Hypotheses Testing.
Topics: IF If statements Else clauses. IF Statement For the conditional expression, evaluating to True or False, the simple IF statement is if : x = 7.
Stefano Vezzoli, CROS NT
Suppose we wanted to estimate the proportion of registered voters who are more enthusiastic about voting in this election compared to other years? Suppose.
How to Avoid the Lies and Damned Lies: Pitfalls of Data Analysis Clay Helberg Special Topics in Marketing Research Dr. Charles Trappey Summarized by Kevin.
Lecture PowerPoint Slides Basic Practice of Statistics 7 th Edition.
1 Chapter 6. Section 6-1 and 6-2. Triola, Elementary Statistics, Eighth Edition. Copyright Addison Wesley Longman M ARIO F. T RIOLA E IGHTH E DITION.
The Scientific Method: Terminology Operational definitions are used to clarify precisely what is meant by each variable Participants or subjects are the.
Reading Report: A unified approach for assessing agreement for continuous and categorical data Yingdong Feng.
Session 1 Probability to confidence intervals By : Allan Chang.
Copyright © 2013, 2009, and 2007, Pearson Education, Inc. Chapter 11 Analyzing the Association Between Categorical Variables Section 11.2 Testing Categorical.
GENERALIZING RESULTS: the role of external validity.
Copyright (C) 2002 Houghton Mifflin Company. All rights reserved. 1 Understandable Statistics Seventh Edition By Brase and Brase Prepared by: Lynn Smith.
Ch 8 Estimating with Confidence 8.1: Confidence Intervals.
13.2 – Find Probabilities Using Permutations A permutation is an arrangement of objects in which order is important. For instance, the 6 possible permutations.
Warm Up Graph each inequality. 1. x ≥ ≤ x ≤ 6 3. x
More on regression Petter Mostad More on indicator variables If an independent variable is an indicator variable, cases where it is 1 will.
10.1 – Estimating with Confidence. Recall: The Law of Large Numbers says the sample mean from a large SRS will be close to the unknown population mean.
T tests comparing two means t tests comparing two means.
Hypothesis Testing. Statistical Inference – dealing with parameter and model uncertainty  Confidence Intervals (credible intervals)  Hypothesis Tests.
©2012 Prentice Hall Business Publishing, Auditing 14/e, Arens/Elder/Beasley Audit Sampling for Tests of Details of Balances Chapter 17.
Inference: Conclusion with Confidence
Chapter 4 Basic Estimation Techniques
Elementary Probability Theory
Process discovery Sander Leemans.
ECO 173 Chapter 10: Introduction to Estimation Lecture 5a
Audit Sampling for Tests of Details of Balances
Angle CAB + Angle ACB + Angle ABC = 180 degrees
ECO 173 Chapter 10: Introduction to Estimation Lecture 5a
Hypothesis Tests for a Population Mean in Practice
Quote of the Day The mathematician may be compared to a designer of garments, who is utterly oblivious of the creatures whom his garments may fit. -Tobias.
Ch8: Sorting in Linear Time Ming-Te Chi
P-value Approach for Test Conclusion
CHAPTER 14: Confidence Intervals The Basics
Chi Square Two-way Tables
A New Product Growth Model for Consumer Durables
Statistical Inference
Analyzing the Association Between Categorical Variables
Chapter 7: Statistical Issues in Research planning and Evaluation
Chapter 8: Estimating With Confidence
Path Analysis for Exploring EBM Science Frameworks
Lesson 1-1 Point, Line, Plane
Enumerating all Permutations
Presentation transcript:

Process Model Realism Measuring Implicit Realism 8/09/2014dr. Benoît Depaire

Research “Trigger” 8/09/2014dr. Benoît Depaire Number of possible execution paths explode with AND-construct with n activities in parallel: X = n! nX

Research “Trigger” 8/09/2014dr. Benoît Depaire ABC ACB BAC BCA CAB CBA

Research “Trigger” 8/09/2014dr. Benoît Depaire ABC ACB CAB CBA

Research “Trigger” 8/09/2014dr. Benoît Depaire ABC ACB CAB CBA R1: All activities must occur

Research “Trigger” 8/09/2014dr. Benoît Depaire ABC ACB CAB CBA R1: All activities must occur R2: A must occur before B, unless we start with C

Research Trigger  When are we generalizing too much?  Context: Capture the true underlying process  Is the discovered model realistic? 8/09/2014dr. Benoît Depaire

Process Realism  Explicit Realism: All observed behavior should be in the model  Implicit Realism: Only the realistic unobserved behavior should be in the model 8/09/2014dr. Benoît Depaire

Assumptions  There are no measurement errors in the log  There are no infinite loops possible in the process  The fitness of a discovered model = 1  (All execution paths are equiprobable) 8/09/2014dr. Benoît Depaire

Process Realism  Implicit Realism: Only the realistic unobserved behavior should be in the model. 8/09/2014dr. Benoît Depaire

Process Realism  Implicit Realism: Only the realistic unobserved behavior should be in the model.  Implicit Realism Measure: How confident can we be that the unobserved behavior is realistic? 8/09/2014dr. Benoît Depaire

Implicit Realism Measure  m = number of paths in process M  n = number of paths in log L  x i = frequency of path i in log L  P i = Probability of path i occurring in L  u = # unobserved paths of M in L  T M (L) = statistic to determine u 8/09/2014dr. Benoît Depaire

Implicit Realism Measure  IR(L,M) = P[T M (L) >= u | M, n]  IR(L,M):  Probability that a model M would generate a log L with at least u missing paths (given n)  The lower IR(L,M), the less confident we can be that M actually produced L  because M contains too much unobserved behavior! (for a given n) 8/09/2014dr. Benoît Depaire

Implicit Realism Measure 8/09/2014dr. Benoît Depaire

Empirical Illustration 8/09/2014dr. Benoît Depaire

Assumptions  There are no measurement errors in the log  There are no infinite loops possible in the process  The fitness of a discovered model = 1  (All execution paths are equiprobable) 8/09/2014dr. Benoît Depaire

Conclusions  IR Measure has a very precise and intuitive interpretation  Current IR Measure should be used for evaluation, not comparison! 8/09/2014dr. Benoît Depaire

Process Model Realism Q&A 8/09/2014dr. Benoît Depaire

Implicit Realism  Precision  To what extent does the model NOT contain too much behavior (no underfitting)  Generalization  To what extent does the model NOT contain too little behavior (no overfitting) 8/09/2014dr. Benoît Depaire

Implicit Realism  Precision  To what extent does the model NOT contain too much behavior (no underfitting)  To what extent does the model ONLY contain observed behavior  Generalization  To what extent does the model NOT contain too little behavior (no overfitting)  To what extent does the model contain unobserved behavior 8/09/2014dr. Benoît Depaire