What Management Science (and Prescriptive Analytics) Can Do for You
Management Science (Prescriptive Analytics and Operations Research), while not new, are more actionable than ever. They enable you to make complex resource allocation decisions – where there are thousands of individual decisions and/or uncertainty about each one – to reduce costs or mitigate risks. Here's the Wikipedia description of Management Science.
With Management Science, you can get systematic, data- and model-based answers for questions like “how do I allocate my marketing budget over email, online ads, social media and more?” or “how do I manage inventory and reorders to minimize stock-outs and holding cost?”
To do this, you need data, but you also need models of your present or future business situation (marketing plans, manufacturing lines, distribution system, financing practices, etc.).
Data Driven, Human-Created Models
The models you need to apply optimization, simulation, and other management science methods are not like the 'models' used in data science, which are standard forms (logistic regression, neural networks, etc.), 'trained' or fitted to past data by machine learning methods.
Building these models requires a person with knowledge of your business -- how a specific function works -- modeling software capable of expressing your business situation in quantitative terms, and solver software to apply optimization and simulation algorithms to your models.
Modeling Systems and Languages
How are management science-based models written and tested? There are three main ways:
- Microsoft Excel -- the tool most business analysts already use -- can be a great tool for management science models. Thousands of businesses have built and are using management science models in Excel, and hundreds of thousands of MBA students have learned management science methods using Excel -- some of them are likely in your organization!
- Algebraic Modeling Languages are specifically designed for management science models. Most such languages are 'purpose-built', such as AMPL, GAMS or AIMMS for optimization, or Arena, Simio, or ExtendSim for simulation.
- SDKs / Object Libraries are popular for building management science models in a programming language. Most vendors of solver software offer an object library for use with their solvers.
Frontline offers a comprehensive platform for both data science (predictive analytics) and management science (prescriptive analytics) that's easy to learn and use in Microsoft Excel, our RASON® modeling language, or a programming language (Solver SDK and XLMiner SDK). In each form, we offer both modeling software and solver software. With Frontline's platform, it's easy to move models between Excel, RASON, and a programming language, and for people working with these tools to communicate and leverage each others' expertise.