O problema com a simulação de sistemas complexos

The problem with simulating complex systems

Simulations are powerful tools that help us predict outcomes, but can they be used when your system is very complex?

Imagem em destaque

explains only 6% of the change in attitudes . Get rid of all the TVs in your house and we're still left with a whopping 94% of unaccounted for variance.

Is there enough data?

COVID was another great example of how simulations can fail dramatically. During the first weeks of the pandemic, some simulations estimated a mortality rate of at least 10%. One in ten infected people would die. Fortunately, this was not the case, but why did scientists miss the target by such a wide margin?

Because at the time we barely had any data on the disease and all the data we had was heavily biased. The sampling came from hospitals, which most people avoided in the first place because they were crowded. So we took samples from people who had pulmonary complications and were at risk.

I'll never get tired of saying this: a model is only as good as the data it's based on . For complex systems, there are a multitude of issues we have to resolve: unstructured data, biased samples, missing data, and uncalibrated equipment are just some of the most common.

Take three different sociological studies on happiness and you will find huge differences in results. This is to be expected considering that sometimes different sources provide different information. For example, nonprofits often criticize governments for sharing incomplete or distorted data.

Other times, the problem isn't in the data collection, but in the fact that the data isn't there . No one could have simulated a global pandemic, plus an accident in the Suez Canal , plus a surge in cryptocurrency, plus one of the worst droughts in recent history in Taiwan – all factors that have influenced the massive chip shortage and rising chip prices. computers we have tried.

At least we have the data now, but what are the odds of these events happening again at the same time?

Are simulations useless?

I know I'm painting a very bleak picture, but the truth is that simulations are extremely important . For hundreds of years, we have simulated situations through experiments, equations and computer models to try to understand how the world works. And through each chance, we also managed to learn a little more.

This is yet another cautionary tale, a reminder that a simulation is neither witchcraft nor a séance but a simple artificial recreation that reproduces the script you were given. We should never take the results of a simulation at face value. As any data scientist knows, we always have to look beyond the results and see how we achieve them. The method is even more important than the result.

That said, the future is bright for simulations. With the Internet of Things and Big Data, we have grown exponentially in our ability to collect data, opening up space for all types of simulations that we considered impossible a few years ago: warehouses, deliveries, market trends, political action.

Excluding the simplest situations, simulations will never be error-free, but we can continue working to minimize this margin of error. Will we ever be able to accurately simulate complex systems? Clear. It's only a matter of time.

Source: BairesDev

Conteúdo relacionado

Back to blog

Leave a comment

Please note, comments need to be approved before they are published.