2. P7: Analyzing prescriptive analytics techniques with
appropriate examples
What is prescriptive analytics?
It is a branch of business analytics that focuses on choosing the optimum course of action in a given set of
conditions or opportunities. Prescriptive analytics can also show you how to maximize a future opportunity or
minimize a future threat, as well as the consequences of each option. Prescriptive analytics is a type of artificial
intelligence that uses advanced mathematical models and algorithms to deliver an automated answer to a
problem.
They are various types of prescriptive analytics some of which include;
• Operational Prescriptive Analytics
• Financial Prescriptive Analytics
• Strategic Prescriptive Analytics
3. Types of prescriptive analytics
Operational Prescriptive Analytics
Operational prescriptive analytics is mostly used in business settings and it is one of the most common type of
prescriptive analytics since it suited for business driven decision making.
Financial Prescriptive Analytics
Financial Prescriptive Analytics is a type of prescriptive analytics that is often combined with operational
prescriptive analytics in order to make decisions related to business’ finance.
Strategic Prescriptive Analytics
This sort of prescriptive analytics is most commonly implemented in situations when business strategies are
created with a clear awareness of short-term objectives.
4. Prescriptive analytics techniques
1) Simulation
Simulation in prescriptive analytics is a method that is used to perform risk analysis by substituting various
variables randomly in a simulation in order to identify unseen events that might affect the possible outcomes in
various ways. This simulation is often applied in the financial sector of a business organization in order to
mitigate the various possible risks related to the reduction in the value of the investments.
An example of simulation is using redesigns of layouts. That is with simulation, it is possible to represent in
details different parts of the variable. Using this, we can different analysis with the model. A similar example
can be seen in the case of Bob and Ray (Baobab, 2020) which can be seen in the next slide.
5. Simulation example
• This example illustrates two people working on an assembly line. The first person Bob begins assembling the
products where after performing a series of operations he finishes and the assembly line is taken over by Ray
who continues with another set of operations. Once the operations have been finished by both people, there is
a finished product as the output.
• However there is no room to store the product between Bob and Ray. That is if Ray has not finished
assembling the product Bob has, Bob gets stuck until Ray finishes. And likewise if Bob has not finished the
assembling.
• Finally, in their various procedures, Bob and Ray do not always take the same amount of time. They can take
anywhere from 10 to 80 seconds to complete. Bob takes 10 seconds 4 percent of the time, 20 seconds 6
percent, and so on, as seen in the table in the next slide.
7. Simulation example (continuation…)
• It's clear that Bob takes 45.9 seconds on average each product, but Ray takes 46.4 seconds, which is slightly
slower. If we used average values, we may anticipate Ray to set the pace because he is the slowest; an
assembled product would be produced every 46.6 seconds, and Ray would be working all the time while Bob
would barely rest.
• However, this simulation ignores the fact that Bob will occasionally take close to 80 seconds and Ray will
take close to 10 seconds, causing Ray to be idle, and vice versa.
• Ray will occasionally take a long time and Bob will take a short time, causing Bob to be blocked. Because
there is no buffer, the production capacity is lower than if Bob and Ray took the same amount of time on
average.
8. Prescriptive analytics techniques (continuation…)
2) Optimization
The process of optimization entails creating a mathematical model with variables and equations whose
resolution allows for the discovery of the optimum solution to a problem. Optimization has been used to solve a
wide range of problems in a variety of industries. From more traditional examples like determining where and
how many buildings to build, through production planning and the construction of effective last-mile supply
routes.
There are also uses that are less well-known but nevertheless very useful, such as the movement of sensors that
examine tissues or samples from patient extractions.
9. Optimization (continuation…)
When information is admitted to be known, optimization models are extremely useful since they will prescribe
the best solution. Uncertainty can also be explicitly incorporated into optimization models. The influence on the
solution given by the optimization model can then be evaluated with all realism using a simulation model, which
can then be used to explore alternatives surrounding it.
10. Conclusion
Finally, optimization and simulation are powerful tools that can assist an analyst in prescribing a remedy based
on the results and outcomes of many variables. Other methods, such as forecasting techniques, can be used in
conjunction with the aforementioned. A forecast model, for example, could aid in reducing uncertainty and
providing more trustworthy data to our simulation and optimization models.