A New Algorithm Makes It Faster to Find the Shortest Paths

Original version Of This is the story Present Quanta magazineThe

If you want to solve a complex problem, it often helps to be organized. For example, the problem is broken into pieces, pieces of pieces in pieces of pieces, pieces of pieces of pieces to pieces. You cut into pieces of pieces of pieces of pieces of pieces of pieces of pieces of pieces of pieces to pieces of pieces of pieces to pieces of pieces of pieces of pieces to pieces of pieces of pieces to pieces of pieces in pieces of pieces cut into pieces of pieces of pieces, cut into pieces of pieces of pieces of pieces of pieces of pieces to pieces of pieces to pieces of pieces to pieces cut to pieces of pieces of pieces of pieces to pieces of pieces to pieces cut to pieces of pieces of pieces of pieces to pieces of pieces be broken into pieces of pieces of pieces of pieces to pieces cut into pieces by pieces of pieces of pieces of pieces to pieces of pieces to pieces of pieces to pieces of pieces to pieces of pieces to pieces cut into pieces of pieces of pieces of pieces to pieces of pieces to pieces cut in pieces of pieces cut into pieces of pieces of pieces of pieces to pieces of pieces to pieces cut in pieces of pieces to pieces of pieces, smumpd to pieces.

This dilemma is especially relevant to one of the iconic problems of computer science: looking for the shortest path at each point from a certain starting point on the network. This is like a soup-up version of the problem that you need to solve when you move away: Learn the best route from your new home, the gym and the supermarket.

“Shortest paths are a beautiful problem that may be related to anyone in the world,” said MickelComputer scientist at the University of Copenhagen.

Consistently, the shortest path to the nearest destination should be the easiest to find. So if you want to design the fastest potential algorithm for the short-blood problem, the closest point, then the next Closast and something else, seem to be justified to start it. However, to do this, you have to determine which point is closest to the time. You will pick the points at the distance. There is a basic speed limit for any algorithm that follows this method: No one can go faster than you take time to pick.

Forty years ago, researchers ran against this “sorting barrier” by designing brief algorithms. Now, has created a team of researchers A new algorithm that breaks itThe It doesn’t pick it up and it runs faster than any algorithm.

“Writers were sad to think that this barrier could break this barrier, said Robert TarjanComputer scientist at the University of Princeton. “This is an amazing result.”

Border of knowledge

To analyze mathematically short-path issues, researchers use the networks or nodes of the language-points of the graph, connected by the line. Each of the nodes is labeled with a number called its weight, which can present the length of that section or the time required to overcome it. Any two nodes usually have many routes and the shortest is the one whose weight adds to the smallest number. Giving a graph and a specific “source” node, the goal of an algorithm is to look for the shortest path to each node.

The The most famous short-path algorithm, Invented The pioneer computer scientist Edsgar is in 1956 by Edsgar Disease, starting from the source and working on the external step by step. This is an effective approach, because knowing the shortest path of the nearest nodes can help you find the shortest paths between the more remote. But the end result is a sort of shortest paths, so the selected barrier sets a basic limit on how fast the algorithm can run.