The portfolio features of Logical Decisions® let you select a complete set of alternatives instead of a single alternative based on the Logical Decisions® ranking results.

LDW’s portfolio features are designed for situations where you want to select a set of alternatives based on budgetary and other constraints. A common application where this situation comes up is in selecting a set (or “portfolio”) of research and development (R&D) projects to fund based on a limited total budget. When selecting the projects, you want to pick those projects that have the best expected performance relative to their price. That is, you want to get the most “bang for the buck.”

The Logical Decisions® portfolio features use the ranking results from a Logical Decisions® analysis for the “bang” part of the equation. It then lets you add cost and other information (the “buck” part of the equation) to identify the portfolio of alternatives that best meet your needs.

View Flyer

More details about the Logical Decisions® portfolio features (LDP)

How LDP Enhances Logical Decisions

Standard Logical Decisions® is designed to help you select a single alternative from a set of possibilities.  The main result of a Logical Decisions® analysis is a utility score for each alternative that summarizes its overall desirability.  The assumption is that you will evaluate your alternatives and then select the one alternative that best meets your objectives – the one that has the highest overall utility.

But, suppose your problem is slightly different.  Suppose that instead of just selecting one alternative you want to select several alternatives.  There are many ways you could go about this.  Four possible approaches are

• Pick the highest ranking alternatives first
• Pick the highest benefit/cost ratio alternatives first
• Use Logical Decisions to rank sets of alternatives, and
• Use an optimization tool to select a set of alternatives

Pick the highest ranking alternatives first. You could use Logical Decisions’ ranking results and pick the first alternative from the list, then the second, then the third and so on until you ran out of budget money.  This approach clearly works, but is it the best?  There are two big problems with this method.  First, it might not give you the most “bang for the buck”, and second, you may have other constraints on your selection that make it unacceptable to select certain combinations of alternatives.

Pick the highest benefit/cost ratio alternatives first.  A second approach is to compute the benefit/cost ratio for each alternative and then to pick the alternatives with the highest ratios until the budget runs out.  This approach is a little better than the first, but there is still no guarantee that you will get the most possible benefit for your budget or that you will meet your other programmatic constraints. You can do benefit/cost analysis using LDP.

Use Logical Decisions to rank sets of alternatives.   A third approach is to use Logical Decisions® to directly rank all of the possible sets of alternatives.  Ideally this is the best approach, since with a good Logical Decisions model you can identify the set of alternatives that best meets your budgetary, programmatic and other goals.  The biggest problem with this approach is that the number of possible sets of alternatives increases exponentially as the number of alternatives increases.  The number of sets of alternatives can become unwieldy even for relatively small numbers of alternatives.  Another potential problem with this approach is that it can be difficult to compute the levels for sets of alternatives based on the levels for the individual alternatives.  However, this method is the preferred approach in the case where selecting one alternative changes the desirability of other alternatives.

Use an optimization tool to select a set of alternatives. The final approach is the one used in LDP.  LDP takes the overall scores of the alternatives from a Logical Decisions® model and uses an optimization method to find the set of alternatives with the highest combined score that also meets the budget.  The optimization tool can quickly search through thousands of possible sets of alternatives and find the one set in all the possibilities that is the best.  Another benefit of this approach is that you can add other programmatic constraints that ensure that your mix of alternatives meets other goals such as diversity, equity, and acceptable use of resources.

The LDP features

LDP is a set of Logical Decisions® features that use the output from a Logical Decisions model as a starting point.  You enhance the Logical Decisions model by adding budget and other constraints.  LDP then performs an optimization that finds the set of alternatives that has the maximum total benefit while meeting your budgetary and other constraints.  LDP can perform either a benefit cost analysis or a true optimization using 0-1 integer programming – a method from the field of operations research that can find the very best set of alternatives that meet the constraints.

LDP is tightly integrated with the standard Logical Decisions® features.  LDP uses the overall benefit (utility) scores for the alternatives from Logical Decisions as the basis for its optimization.  It also uses Logical Decisions measures to describe the the cost and resource usage information for each alternative.

LDP can include cost and budget information for many different budget periods or categories.  You can define a budget and corresponding cost for each alternative for each period or category.  The optimization engine will make sure that the selected alternatives cost less than the allowed budget for every category that you define.

You can experiment with the effects of different budgetary and other constraints by defining different “scenarios.”  Each scenario has its own budgets, optimization goal and programmatic constraints.

LDP allows a variety of programmatic constraints.

• You can force any alternative into or out of the final selection.
• You can add if-then constraints that ensure that alternative A is only chosen if alternative B is also chosen.
• You can define groups of alternatives and force all of the alternatives in the group into or out of the selection, force either one or none of the alternatives to be selected, all or none of the alternatives to be selected or exactly one of the alternatives to be selected.
• You can add resource constraints such as available personnel and ensure that your selected alternatives don’t use more than the available amount of the resource.
• Finally, you can add allocation constraints that specify a minimum percentage of the budget that must be spent on alternatives belonging to a particular group.

All of these constraints can be added in an easy, intuitive way that shields you from the complexities of the mathematical model structure.

Sensitivity analysis and the efficient frontier

LDP lets you perform sensitivity analysis on your budgets to see how the total benefit increases as the available budget increases.  The result is a true efficient frontier showing the maximum benefit that can be achieved at each budget level while still meeting the programmatic constraints.  Contrast this to other tools that just order the alternatives by their cost/benefit ratios and don’t even look for the alternatives that maximize benefits at each budget level, let alone try and meet the programmatic constraints.

Copyright © 2021 Logical Decisions

 1,156 total views,  2 views today