ITIS 1 MID-TERM
The Modeling Process
-Quantitative Business Modeling
1. Opportunity/ problem recognition (most important step)
-determining its magnitude, defining it precisely.
-problems are flipside of opportunities
-should carefully investigate the situation of concern.
2. Model Formulation (most time consuming steps according to experts)
-transforming the complex opportunity or problem into a mathematical formula
-the simpler the model, easier the manipulations and the solution but the less
representative it will be of the real life problem.
-most important is determining the components and key variables of the model
when influence diagram is drawn.
3 basic components: independent, dependent, and uncontrolled variables.
3. Data Collection
4. Analysis of the Model
-how well the model represents reality
-are they empirically accurate
-representative of system’s behavior under real world situation
5. Implementation and Project Management
-communicating the results of the study and the modeler’s recommendation
-actually applying managerially approved action through the project management
-evaluate the results
• Easier than manipulating real system
• Cost of mistakes is smaller
• Enables compression of time
• Helps managers understand tradeoffs between different approaches to a management
• Enables quick identification and analysis of a very large # of possible situation
• Aids handling large amount of information
• Time consuming, expensive
• Reluctant to accept results
• Obtaining necessary data might be difficult
• Difficult to access uncertainties
• Perception that it actually eliminates the risks and uncertainties
• Results may be used without appropriate validation
o Validation - establishes how well the model represents reality
Types of variables, observations, & their properties
Approaches to data collection: groups or cross-sectional and time series Ishani Sinha
Types of Observations
Definitions of probability:
Mean Variance Standard Deviation
Excel can calculate both – variance of population and variance of sample
Excel always calculates sample – standard deviation of population and standard deviation of
Statistical Displays / Frequency Tables: When are they used?
• Bar chart : ordinal data, arranged in order
• Pareto Chart: nominal data, arranged highest to lowest
• Histogram: numerical data, no space between bars
3 Types of probabilities Ishani Sinha
1. Subjective Probability – individual information and belief
2. Logical Probability – physical or deterministic phenomena
3. Experimental – frequencies based on trials
5 Types of events
1. Independent events occurrence of one event does not affect the probability of
occurrence of the others. (head / tails of a coin)
2. Dependent events occurrence of one event affects the occurrence of the next event.
(drawing a second diamond card from the deck)
3. Mutually exclusive events occurrence of one precludes the occurrence of the others.
(birth of a child -> “it’s a boy” or “it’s a girl”)
4. Collectively exhaustive if one event occurs, then the set of events needs to occur
(when rolling a die, 1, 2, 3, 4, 5, 6 must occur for the event to occur)
5. Complements Law (p,q) the probability of an event not occurring (p) then the
probability of the opposite event to occur is (1-p) (if p=0.6 then q=0.4)
2 Formulas (‘AND’ & ‘OR’)
1. The Multiplication Rule
Joint Probability for dependent events Conditional probability for dependent events