A local optimization approach can be more ambitious in finding the best solution, which translates into a good solution, but not the best. Using a more global optimization approach can take longer and the solution obtained should be more reliable. Global optimization problems are common in organizations, classically referred to as “conflicts between principal agents”. Individuals or departments often prioritize their own immediate needs and interests over what is best for the entire organization, which, in most cases, would be best for everyone.

In some cases, everyone works well together, but they're going down the wrong hill because the organization has the wrong objective. Well-structured incentives and organizational culture can help in this regard. Howard Schultz thoroughly analyzes in “Onward” (O review + notes) how focusing too much on “compensations” (sales in the same store) led Starbucks to lose its way in the mid-2000s. Global optimization is a fairly simple concept to understand, so I won't dwell on the mechanics.

I first learned it in an operations management class, where the professor commented that some linear optimization algorithms (such as the Excel “Solver”) adopt an approach similar to that described by Norman above. So, although I will cite a lot of great books in the recommendations section on the local versus. Global optimization in classic business contexts, such as “disruption”, here I will try to offer some more innovative interpretations. If you're impatient, ask for “Onward” by Howard Schultz (Or review + notes), “The Container Store” by Kip Tindell (review of TCS + notes), “The Everything Store” by Brad Stone (review of TES + notes) or “Made in America” by Sam Walton (review of MA + notes).

Turns out Walton was also a disruptor. One of the most intriguing and unusual applications of the venue vs. The global optimization model that I have encountered relates to social connections. Are you starting to watch local games?.

Here's the global optimization angle? If you want to reach the “global optimal” of being a world-class forecaster (or a thinker in general), you must make local decisions that are not optimistic. How are the people around you likely to perceive you if you start to communicate in the “multifaceted” and nuanced way that Tetlock's eponymous superforecasters did? The sad answer is: “It's not right. On pages 138 to 139 of Superforecasting, Tetlock explains:. In this tutorial, we'll talk about the concepts of local and global optimums in an optimization problem.

First, we'll make an introduction to mathematical optimization. Then, we will define the two terms and briefly present some algorithms for calculating them. First, let's briefly introduce the basic terms of optimization. If the objective function contains more than one global optimal, the optimization problem is called multimodal.

In this case, there are several values of the decision variables that optimize the objective function. In general, finding the local and global optimals of a function is a very important problem, and new algorithms are being proposed all the time. However, to choose an algorithm, we must always first understand the requirements of our optimization problem, since there are algorithms that claim to be fast and lose a little precision and adversely. Local optimization, where the algorithm can get stuck in a local optimal without finding a global optimal.

The presence of the local optimal is an important component of what defines the difficulty of a global optimization problem, since it can be relatively easy to locate a local optimal and relatively difficult to locate the global optimal. Finding an arbitrary local minimum is relatively simple using classic local optimization methods. Global optimization is distinguished from local optimization by its focus on finding the minimum or maximum in the given set, as opposed to finding local minimums or maximums. The global optimal may be the same as the local optimal, in which case it would be more appropriate to refer to the optimization problem as local optimization, rather than global optimization.

. .