Artificial intuition provides a better answer

Assessing the optimal location for car parks is a surprisingly mathematical challenge, a subset of a classic computational complexity problem with much broader applications. A team of data scientists has combined quantitative annealing with a process that attempts to mimic the basic processes of human intuition in a technique that provides much higher resolution accuracy than traditional methods.

This technique is described in a research paper that first appeared online September 30, 2022 in An intelligent, converged network.

The facility location problem, or FLP, is a long-standing challenge within operations research—the application of scientific methods to decision-making and problem-solving by managers in large commercial, public, or military organizations. FLP aims to determine the optimal location and number of facilities in a given area, subject to some limitations

Decision makers in the health care system for example may have to assess where a new hospital should be located. If they place this facility in a location that is difficult for the elderly to access, mortality rates may increase. The limitation here is trying to reduce the mortality rate. But if they put the hospital in an accessible location, real estate costs could eat more out of their budget—another drawback. This may also reduce the ability of healthcare providers to provide the service, which again increases mortality rates.

It might seem like figuring out one or even any number of sweet spots where death rates and spending are at their lowest is mathematically simple. But finding exact solutions for this and other examples of FLP – with many limitations on what to optimize than just distance and cost – is computationally challenging. In fact, the FLP, a combinatorial optimization problem, has been classified by complexity theorists to be in the category of “NP-hard” problems – the hardest. There is no single solution that can be applied to site planning for different situations in different fields.

Optimal parking location is just another example of FLP, and one that is of great interest to city managers wanting to avoid congestion, and thus reduce greenhouse gas emissions. The less time motorists spend searching for a parking space, the less traffic and the less greenhouse gas emissions. In many cities in rapidly urbanizing regions, not least in the developing world, this is a pressing concern.

Conventional approaches used to produce adequate (but not exact) solutions to the parking lot location problem involve various types of algorithms performed by artificial intelligence on classical computers (as opposed to quantum computers). But once the amount of data involved gets too high, the performance of these “classical intelligent algorithms” drops sharply.

“At this point, a person’s intuition can overtake a computer,” said Sumin Wang, a co-author of the paper and a researcher at the Key Laboratory of Specialized Optical Fibers and Optical Access Networks at Shanghai University. But this intuition should not be seen as just “mystical feelings.” There is a strong scientific explanation for where human intuition comes from, and this could inspire us to try to imitate it with computers.”

When an engineer or architect has a feeling that a bridge, building system, or other engineering structure is about to collapse, but cannot directly justify the cause, this may be the result of decades of experience. A cyclist can feel exactly when his bike is about to tip over without being able to explain what he was feeling that allowed him to make such an assessment.

The vast experience and accumulated knowledge can allow an individual to quickly assess the totality of a situation and perceive the truth firsthand without working through a traditional thought process, and allow for quick and effective decision-making despite complex environments.

Those who study human intuition describe what happens in the brain as a rapid, sharp decrease in the “search space”—the term computer scientists use to describe the landscape of possible solutions. Experience and knowledge allow humans to “know” how to selectively pay attention to the most prominent aspects of a problem, and ignore the rest, thus simplifying the necessary calculations.

“Artificial intuition,” which artificially replicates human intuition, is an emerging research area in the field of artificial intelligence. The goal is to develop intuitive reasoning techniques inspired by the human brain—one of the most powerful abilities we possess—that similarly focus on essential data while ignoring unimportant data to narrow the search space.

Using the optimal parking lot location problem, the researchers developed what they call the Selective Attention Mechanism (SAM), which was inspired by human intuition, and combined with quantum annealing (QA).

Discrete QA has received a lot of attention in recent years as a new computational model for solving classic optimization problems. Quality assurance algorithms provide significant improvements in terms of algorithm runtime and solution quality to some difficult NP problems that are poorly solved by classical methods. In optimization problems, one looks for the best possible number of combinations, as a minimum or maximum. And in physics, everything searches for the lowest energy state, from balls rolling down hills to excited electrons returning to their ground state. This means that optimization problems can in essence be recast as energy reduction problems. QA only exploits quantum physics to determine the lowest energy states of the problem, and thus the minimum or maximum of the target characteristic. QA has already been deployed in many applications from traffic optimization to resource scheduling and quantum chemistry.

For the parking optimization problem, the researchers used SAM to reduce the search space and provide guidance for the next search step, and QA to search that space and improve search efficiency.

They applied their concept to a real-world parking experience using real latitude and longitude data from Luohu District in Shenzhen, China. This platform’s open government data included locations with high parking demand, potential parking locations, current parking lots, and their parking capacity. Luohu area covers an area of ​​about 80 square kilometers, which is too large for any classical intelligent algorithm with limited computing resources to directly calculate all the data in the area.

The entire region was first divided into blocks to save computational resources, and then SAM was applied to focus on important data points, which were automatically filtered and optimized. The new location results were then obtained by simulating the QA preference for lower-energy states. The selective POIs were updated in turn based on the results of the facility’s location, and the process was repeated several times until an obvious solution appeared – the location of the proposed new parking lot in the area –

To evaluate their approach, the researchers used a commonly used technique for measuring the resolution accuracy of multi-objective algorithms. Compared to competing approaches, SAM plus QA technology has produced better and more feasible solution sets at a shorter run time.

The researchers now want to take their approach and apply it to other positioning problems and related applications of artificial intuition.

The paper is also available on SciOpen (https://www.sciopen.com/article/10.23919/ICN.2022.0017) via Tsinghua University Press.

###

About Smart and converged networks

Smart and converged networks It is a specialized international magazine focusing on the latest developments in communications technology. The journal is jointly published by Tsinghua University Press and the International Telecommunication Union (ITU), the United Nations specialized agency for information and communication technologies (ICT). Intelligent and converged networks derive their name from the accelerating convergence of various fields of communication technology and the growing influence of artificial intelligence and machine learning.

About SciOpen

SciOpen is a professional open-access resource for discovering scientific and technical content published by Tsinghua University Press and its publishing partners, providing the scientific publishing community with innovative technology and market-leading capabilities. SciOpen offers comprehensive services across manuscript submission, peer review, content hosting, analytics, identity management, and expert advice to ensure the development of each journal by offering a range of options across all functions such as journal planning, production services, editorial services, marketing and promotion, online jobs , etc. By digitizing the publishing process, SciOpen expands reach, deepens impact, and accelerates the exchange of ideas.

Not giving an opinion: AAAS and EurekAlert! Not responsible for the accuracy of the newsletters sent on EurekAlert! Through contributing organizations or for using any information through the EurekAlert system.

Leave a Comment