Chapter 7

• Introduction to Models in Decision Support Systems

• Models are an integral part of Decision Support Systems (DSS),
helping simulate real-world problems and providing a framework for
analyzing potential outcomes. In this lecture, we will explore the
different types of models used in DSS, dive into discrete-event
simulation models, understand the importance of random and
pseudo-random numbers, and look at basic simulation models. These
components are essential for decision-makers aiming to model
uncertainty, predict outcomes, and optimize processes.
• Types of Models in DSS
• Models in DSS can be categorized based on the type of problem they address and the decision-making context. The major types include:
• a. Mathematical Models:
• These models use mathematical equations to represent relationships among variables.
• Types of Mathematical Models:
• Linear Models: Used when the relationship between variables is linear. Example: Linear Programming.
• Non-Linear Models: For complex relationships where the relationship between variables is not linear. Example: Non-linear regression.
• Optimization Models: Aim to maximize or minimize a particular function (profit, cost, etc.).
• Stochastic Models: Incorporate random variables and are used for problems involving uncertainty.
• b. Simulation Models:
• Used to simulate the behavior of a system over time.
• These are ideal when real-world experimentation is too costly or impractical.
• Simulation models can either be deterministic (no randomness) or stochastic (includes randomness).
• Examples: Inventory management systems, supply chain simulations.
• c. Expert Systems:
• These are rule-based systems that capture human expertise and apply it to decision-making.
• Often used in knowledge-driven DSS.
• d. Statistical Models:
• Utilize statistical techniques to make predictions or inferences from data.
• Examples: Regression analysis, hypothesis testing.
• Discrete-Event Simulation (DES) Models
• Discrete-Event Simulation (DES) models are used to represent systems where changes happen at distinct points in
time. This is ideal for systems where events (e.g., arrival of customers, machine breakdowns) occur intermittently.
• Key Characteristics:
• Events: These are occurrences that change the state of the system (e.g., a customer arriving at a bank or a
machine starting production).
• Entities: Objects that experience the events (e.g., customers, products).
• State Variables: Variables that define the system at any point in time (e.g., number of customers in line, machine
status).
• Applications of DES:
• Manufacturing Systems: DES models can simulate production lines, identify bottlenecks, and optimize throughput.
• Queue Management: Used to model queues in banks, call centers, hospitals to improve service and reduce wait
times.
• Logistics and Supply Chains: Simulates inventory levels, order processing times, and transportation to optimize
supply chain performance.
• Random Numbers, Pseudo-Random Numbers, and Statistical Distributions
• Randomness plays a crucial role in DSS, especially in simulation models, where systems need to account for uncertainty.
• a. Random Numbers:
• True Random Numbers are generated from physical phenomena (e.g., radioactive decay, atmospheric noise).
• Often used when an unpredictable or completely random sequence is needed.
• b. Pseudo-Random Numbers:
• Pseudo-Random Numbers (PRN) are generated using algorithms and appear random but are actually deterministic.
• They are essential in simulations where repeated testing with consistent random sequences is necessary.
• PRNs are widely used due to their ease of generation and computational efficiency.
• c. Statistical Distributions:
• Simulation models often use statistical distributions to represent real-world variability.
• Key Distributions include:
• Normal Distribution: Models many natural phenomena (e.g., heights, test scores).
• Exponential Distribution: Commonly used for modeling time between events (e.g., time between customer arrivals).
• Poisson Distribution: Used for modeling the number of events occurring within a fixed interval of time.
• Uniform Distribution: Every number in a range has an equal probability of occurring.
• Understanding and using these distributions correctly is essential to accurately simulate systems and their behavior under
uncertainty.
• Basic Simulation Models
• Simulation models are powerful tools for analyzing complex systems without physically testing them. Basic simulation models are foundational to
more advanced techniques.
• a. Monte Carlo Simulation:
• A widely used technique that relies on repeated random sampling to calculate results.
• It’s used for risk analysis and decision-making in uncertain environments.
• Applications include finance (e.g., risk modeling), engineering, and project management.
• b. Markov Models:
• These are stochastic models that represent systems where transitions between states occur according to probabilistic rules.
• Used in situations where outcomes are dependent on the current state but not on the sequence of events that preceded it (i.e., memoryless
property).
• Commonly applied in economics, finance, and healthcare decision-making.
• c. Queueing Models:
• These models simulate queues or waiting lines.
• They are often used to analyze systems where resources need to be allocated efficiently (e.g., customer service, telecommunications).
• Important parameters include arrival rates, service rates, and the number of servers.
• d. Inventory Models:
• Used to determine optimal order quantities and inventory levels to minimize costs associated with holding, ordering, and stockouts.
• Popular techniques include the Economic Order Quantity (EOQ) model and the Newsvendor model.