My favorite hard issue is the traveling salesperson issue. When you are presented with a list of cities that are in the same region, it is asked what would be the best way to cover every one of them and return to the town of origin. To find concrete solutions to real-world problems, scientists employ approximate algorithms, techniques that do not solve the issues precisely but are close enough to provide useful. The most efficient algorithm, which was developed in the year 1976, was able to guarantee that the answers it offered were not more than 50% of the correct answer.
I’m a specialist in approximate algorithms as a researcher in the field of computer science. My coworkers Anna Karlin and Shayan Oveis Gharan and I have discovered an approach to surpass the 50% mark; however, only a little. We were able to prove that a particular approximation algorithm can break through the long-standing obstacle and will allow for further important advancements.
This is more important than simply making plans for routes. All of these problems could be encoded into the travel salesperson problem and vice versa: solve one, and you’ve solved all. These tough problems are the same computational gremlin with different hats.
The best way isn’t easy to locate.
The question is often framed in terms of “find the shortest route.” However, the most efficient solution may be determined by a variety of variables that are actually in use, like costs, time, and distance.
To understand the reasons for this issue to solve, consider the following scenario: If someone gives you 100 cities, along with the cost of train, plane, and bus tickets for each of them. Can you find the most affordable route which covers all of them?
Think about the vast possibilities of routes. If there are 100 cities that you would like to visit, the total number of possible ways to go is 100 factorial, that is 100 99 x 98 … 1. This is more than the number of atoms present in the universe.
This set of lines and dots is the shortest-selling problem tour, which covers 1,000 points. William Cook et al., CC BY-ND
Just enough for the job.
The fact that these issues aren’t easy isn’t a reason to hinder them from appearing within the world of work. In addition to finding routes for travel salespeople (or nowadays delivery trucks), The problem of the traveling salesperson is applicable in a wide range of areas, including the mapping of genes to creating circuit boards.
To address real-world scenarios of this issue, professionals follow the same method that people have been doing for thousands of years: find solutions that may not be ideal but are adequate. It’s fine that a salesperson chooses a route that’s only more than what it should be. Nobody cares whether a circuit board takes only a fraction of a second longer to make or an Uber will take a couple of minutes longer to get its customers home.