MATHEMATICAL MODELS

This diagram categorizes models based on their implementation methods into two primary
categories: Analytical and Algorithmic.

1. Analytical:

 Algebraic: Uses exact formulas to solve models, often expressed as equations with
closed-form solutions.
 Approximate: Provides solutions that are not exact but close enough to be useful, often
used when exact solutions are too complex or impossible. Example weather patterns,
Economic models etc.

2. Algorithmic:

 Numeric: Involves solving mathematical models using numerical methods, This is

achieved, for example, with the transition from continuous mathematical equations to
discrete ones and then developing the computational algorithm to solve the final set of
equations. Example changes of temperature.
 Simulation: Replicates the behavior of a system over time, often using computational
techniques to model dynamic and complex processes.

TYPES OF SIMULATION MODELS

Simulation models can be classified as being deterministic or stochastic, static or dynamic and
discrete or continuous.
So, Are the Listed Models Analytical or Simulation Models?

 Simulation Models:
o Discrete-Event Simulation, Dynamic models, and Monte Carlo Simulation
(static + stochastic) are clearly simulation models, as they focus on imitating
processes and dynamics over time.
 Mathematical Models Embedded in Simulations:
o Deterministic Models, Stochastic Models, and Continuous Simulation Models
are built on mathematical models (such as probability theory, differential
equations, or algebraic models).
o These become simulation models when used in computational experiments or
simulations over time, often involving numerous iterations or variations.

Thus, many of these models are hybrid in the sense that they are based on mathematical models
but are implemented and solved using simulation techniques when analytic solutions are not
practical.

Stochastic Models vs Deterministic Models

Stochastic Models
These models involve randomness and uncertainty. The results are expressed in terms of
probability distributions, and running the model multiple times with the same initial conditions
may yield different outcomes.

 Example: A queuing model in a call center where customer arrival times are random and
follow a probability distribution.
 Applications: Stock market simulation, weather forecasting, biological population growth
models.
Deterministic Models
 In deterministic models, the outcome is fully determined by the initial conditions and the
underlying equations. There is no randomness or uncertainty in the model's behavior.
Given the same inputs, the model will always produce the same results.
 Examples: Linear programming models, certain types of optimization models, or
physical models like Newton’s laws of motion when applied in a controlled environment.
 For example, a simulation of a factory production line that models machine outputs
deterministically (i.e., based on a fixed process and not random variation) could be
considered a deterministic simulation.

Static vs. Dynamic Models

Static models
Model where time is not a significant variable
Examples
 Determine the probability of a winning solitaire hand
 Static + stochastic = Monte Carlo simulation
 Statistical sampling to develop approximate solutions to numerical problems
A simple example of a static model in mathematics is the solution to a linear equation in two
variables. Consider the linear equation:
y=2x+3
This equation represents a straight line on a Cartesian coordinate system.
Dynamic models
Model focusing on the evolution of the system under investigation over time
A simple and classic example of a dynamic mathematical model is the basic model of population
growth. The most fundamental version of this model uses a differential equation to describe how
a population changes over time, considering a constant rate of growth. This model is known as
the exponential growth model.
Here's the mathematical representation:

The solution to this differential equation is an exponential function

In this model, the population grows continuously and exponentially over time. It's a simplistic
model as it assumes unlimited resources and no constraints on growth, which is rarely the case in
real-world scenarios. However, it serves as a fundamental starting point for understanding more
complex dynamics in population biology and ecology.
Continuous vs. Discrete
A Discrete Model
The discrete State of the system is viewed as changing at discrete points in time
A simple example of a discrete model can be found in the concept of a queue, such as the line of
customers waiting at a bank. In this model, the customers and their arrival and service times are
treated as discrete events.

Queue at a Bank: A Discrete Model

Consider a queue at a bank where customers wait to be served by a teller. Here's how this
scenario can be modeled discretely:

 Discrete Entities: Each customer is a discrete entity. They arrive at the bank at specific,
countable moments in time.
 Discrete Events: The arrival of a customer is a discrete event. Similarly, the beginning
and end of service for each customer are discrete events.

In summary, the queue at a bank is an excellent example of a discrete model, where the entities
(customers) and events (arrivals and departures) are distinct and countable. This model helps in
understanding and managing systems where events occur at separate points in time.

Continuous
 The state of the system is viewed as changing continuously across time
 System typically described by a set of differential equations
A simple example of a continuous model is the monitoring of temperature over time. In this
model, temperature is recorded as a continuous variable that changes smoothly and steadily,
without discrete jumps or intervals.

MODELS TO BE COVERED: MONTECARLO. DISCRETE AND CONTINUOUS

STEPS IN A SIMULATION STUDY

  Problem formation Problem formation
 Model construction
 Data Collection
 Model programming
 Validation
 Design of experiment
 Simulation run and analysis
 Documentation
 Implementation

A simulation study typically follows a structured set of steps to ensure that the system being
modeled is understood, the simulation is built correctly, and the results are analyzed and
interpreted meaningfully. Below are the common steps in conducting a simulation study:

1. Problem Definition

 Objective: Clearly define the problem or system to be studied. Identify the goals of the
simulation, the key questions to be answered, and the scope of the study.
 Key Activities:
o Understand the system or process to be modeled.
o Identify the objectives and metrics (e.g., efficiency, cost, time).
o Set the boundaries and assumptions for the simulation.
 Example: In a hospital setting, the goal could be to reduce patient waiting times by
optimizing the flow of patients through different departments.

2. System Conceptualization (Model Construction)

 Objective: Develop a conceptual model of the system, which is a simplified

representation of the actual system. This step includes identifying the main components
and their interactions.
 Key Activities:
o Define the system components (e.g., machines, people, or agents).
o Understand the relationships and flows between components.
o Make assumptions about system behavior if necessary.
 Example: In the hospital scenario, conceptualize departments (emergency, surgery,
outpatient) and patient flow between them.

3. Data Collection and Analysis

 Objective: Gather the necessary data to accurately model the system. This includes
collecting both qualitative and quantitative data.
  Key Activities:
o Identify input variables (e.g., arrival times, processing times, service rates).
o Collect historical or experimental data.
o Analyze the data to understand distributions, correlations, and trends.
 Example: Collect data on patient arrival times, the time spent in each department, and the
number of available staff or resources.

4. Model Formulation

 Objective: Translate the conceptual model into a formal model, either mathematical or
rule-based, depending on the simulation type.
 Key Activities:
o Develop the equations, logic, or rules that govern the system’s behavior.
o Choose the type of simulation model (e.g., discrete-event, dynamic or any other).
o Use simulation software or programming languages to build the model.
 Example: Build a discrete-event simulation that models patient arrivals, service rates,
and queueing at different hospital departments.

5. Model Verification

 Objective: Ensure that the simulation model is built correctly and behaves as intended.
Verification checks if the model's implementation is error-free and performs as expected.
 Key Activities:
o Debug the model to find and correct coding or logical errors.
o Check if the model is consistent with the conceptual model.
o Test with simple scenarios to ensure the model behaves as expected.
 Example: Run a few test cases with controlled input (e.g., a fixed number of patients) to
ensure the model behaves correctly.

6. Model Validation

 Objective: Ensure that the simulation model accurately represents the real-world system
it is intended to simulate. Validation compares the model's outputs with real-world data
or expert opinions.
 Key Activities:
o Compare model results to actual system behavior or historical data.
o Involve domain experts to check if the model makes sense.
o Perform sensitivity analysis to check the model’s robustness.
 Example: Compare the simulated patient waiting times with actual historical waiting
times to validate the model.
7. Experimental Design

 Objective: Design experiments to test different scenarios and identify the factors that
influence the system's performance.
 Key Activities:
o Define the input variables and scenarios to be tested.
o Decide on the number of simulations runs for each scenario.
o Consider stochastic variability (randomness) in multiple runs.
 Example: Run the simulation for different staffing levels, patient arrival rates, and
department layouts to see how these factors impact waiting times.

8. Simulation Execution

 Objective: Run the simulation model to generate output data for analysis.
 Key Activities:
o Perform multiple simulation runs for each scenario.
o Collect output data (e.g., waiting times, system utilization).
o Ensure enough simulation runs are performed to account for variability.
 Example: Simulate patient flows over a one-month period with varying arrival rates and
staffing levels, collecting data on key performance indicators (KPIs).

9. Output Analysis

 Objective: Analyze the results of the simulation to gain insights into system behavior and
performance.
 Key Activities:
o Analyze output data using statistical methods to identify trends, averages, and
distributions.
o Use visualization tools (e.g., graphs, charts) to present the results.
o Identify which input variables have the most significant impact on the system.
 Example: Analyze patient waiting times, department utilization, and bottlenecks to
identify where improvements can be made.

10. Implementation of Results

 Objective: Use the insights gained from the simulation to make informed decisions or
recommendations for improving the real-world system.
 Key Activities:
o Provide recommendations based on the simulation outcomes (e.g., increase
staffing, change patient flow).
o Discuss the findings with stakeholders.
 o Implement changes in the real-world system, if appropriate.
 Example: Based on the simulation, the hospital may increase staff during peak hours to
reduce patient waiting times.

11. Documentation and Reporting

 Objective: Document the entire simulation study and communicate the findings and
recommendations to stakeholders.
 Key Activities:
o Document the assumptions, data sources, model formulation, and validation steps.
o Prepare a report with a clear presentation of the results and recommendations.
o Include visualizations and summaries for clarity.
 Example: A final report on patient flow in the hospital, including simulation findings,
bottlenecks, and proposed solutions.

12. Model Maintenance and Updates

 Objective: Update the model as new data becomes available or the system changes over
time. This step ensures that the simulation remains relevant and useful for future
decisions.
 Key Activities:
o Revise the model to reflect changes in the system or environment.
o Incorporate new data or updated assumptions.
o Rerun the simulation as needed to provide updated recommendations.
 Example: If the hospital adds a new department or significantly changes staffing
policies, update the simulation to reflect these changes.

Conclusion:

By following these steps, a simulation study can effectively model complex systems, test various
scenarios, and provide valuable insights for decision-making. Each step is crucial in ensuring the
simulation model is accurate, reliable, and provides useful results for real-world applications.