**Frontline Solvers** support several key analytics methods. This page provides a brief overview of these methods, and how they can help you gain insights and make better decisions. Click the “learn more” link at the end of each section for more about each analytic method.

## Forecasting and Data Mining (Better Predictions from Data)

Forecasting and data mining enable you to make predictions based on data, rather than "gut feel." **Data visualization** tools summarize thousands of data points visually, and help you see patterns and gain insights right away. Statistical **forecasting **tools can extrapolate a "time series" (such as past sales figures, or exchange rate quotes) to predict the next value. **Data mining** tools sift through past observations of many possible variables to find the best explanatory variables, and use them to either **classify** a new case (such as "good" or "bad" credit risk) or **predict** an outcome variable (such as a home price).

The software tools use statistical and machine learning methods to "fit the parameters" of a predictive model, but you'll use human intelligence and automated aids to select and (often) transform the data to be analyzed, "train" the model on some data and evaluate its performance on other data, and compare results from the different methods to arrive at a good model you can use in practice.

Learn more about **Forecasting and Data Mining** ➡

## Conventional Optimization (Better Decisions about Resources)

**Conventional optimization** enables you to *optimally allocate* scarce resources: raw materials, machines, people, money, or anything else in limited supply. An optimization Solver can search through millions of combinations of ways to allocate resources, and find the best way, where you define what is “best.” You don’t have to write a search program – you just create a model that defines:

- An
*objective*formula you want to maximize (such as profit) or minimize (such as cost). - What
*decisions*(how much of what resource to allocate for some purpose) you can make. - What
*constraints*(such as supply limits or operating rates) the solution must satisfy.

The solution gives you best values for all the decisions. **Conventional optimization** is the right method in situations with little *uncertainty*: you know with fairly high confidence what your costs, limits, rates, and other parameters of the model will be, and you are focused on making a decision “here and now”.

Learn more about **Optimization** ➡

## Risk Analysis Using Monte Carlo Simulation (Better Understanding of Risk)

**Monte Carlo Simulation** helps you understand and quantify the range of potential outcomes in situations where there is significant *uncertainty* – factors such as weather, stock prices, or customer or competitor actions you don’t control that can impact your results – while holding constant the key decisions you do control. You simply create a *model* of the business situation where you:

- Identify the
*uncertain variables*(inputs) that affect your results. - For each uncertain variable, choose a
*distribution*that estimates the range of values it can take. - Identify the
*uncertain functions*(calculated results) most important to you.

A simulation engine or Solver runs thousands of “what-if” scenarios, each time sampling possible values for uncertain variables and computing results, and then summarizes the results for you in charts, graphs and statistics. You can see the *full range* of outcomes (which is often surprising) and quickly quantify the likelihood of acceptable or unacceptable results.

Learn more about **Monte Carlo Simulation** ➡

## Simulation Optimization (Better Decisions in Risky Situations)

Optimization helps you make better choices when you have all the data, and simulation helps you understand the possible outcomes when you don’t. Frontline Solvers enable you to *combine* these analytic methods, so you can make better choices for resources you do control, taking into account the range of potential outcomes for factors you don’t control.

This method, called **simulation optimization**, helps you make better resource allocation choices “here and now,” in situations with uncertainty. You simply create a model that includes:

*Decision variables*for resources you do control, just as in a conventional optimization model*Uncertain variables*for factors you don’t control, just as in a Monte Carlo simulation model- An
*objective*to maximize or minimize, that may depend on decision and uncertain variables *Constraints*to satisfy, that may also depend on decision and uncertain variables

A Solver using **simulation optimization** can search through thousands of ways to allocate resources and, for each one, through thousands of possible future outcomes, and seek the *best* set of choices. The solution gives you a much better picture of the decisions you should make, given what you know for sure and the range of outcomes for factors that you don’t know for sure.

Learn more about **Simulation Optimization** ➡

## Stochastic Optimization (Better Plans for a Risky Future)

Simulation optimization has limitations when you are planning for the future: It only allows you to make decisions “here and now,” then accept the uncertain outcome. In many real-world situations, once an outcome occurs, you can take corrective or compensating actions. These are “wait and see” decisions – and when you can make *these* decisions, it usually *changes* the best choice of decisions “here and now”.

**Stochastic optimization** captures both “here and now” decisions and “wait and see” (or recourse) decisions in your model. Hence, it enables you to find the *optimal plan* for a *risky future* and can often result in different, better, decisions being identified than what will be identified using only Simulation Optimization. A **stochastic optimization** model isn’t much harder to create than a simulation optimization model. It will include:

*Uncertain variables*for factors you don’t control, just in a Monte Carlo simulation model*Conventional*(“here and now”) decision variables for choices you must make today, before the uncertain outcomes are known*Recourse*(“wait and see”) decision variables for choices you can make later, when the uncertain outcomes are known- An
*objective*to maximize or minimize, that may depend on decision and uncertain variables *Constraints*to satisfy, that may also depend on decision and uncertain variables

The solution you get from stochastic optimization includes both *single* best values for “here and now” decisions, and multiple best values for “wait and see” decision. You get a complete plan, with guides for action in *multiple* future scenarios. This powerful capability is only available in Risk Solver Platform.

Learn more about **Stochastic Optimization** ➡

The more you know, the more you will appreciate how **Frontline Solvers,** more than any competing alternative, empower you to use the latest techniques and fastest Solvers to help you make the best decisions with the least work.

And because we’ve built a solid upgrade path, whether it is adding new capabilities to solve *new types* of problems, or using our full range of plug-in **Solver Engines** to solve even the largest problems, you know your investment today will still be paying dividends even as your needs change and grow.