Experienced stainless steel work from China
The cover of the British “Nature” magazine, titled “Matrix Game”, published the latest discovery of the artificial intelligence (AI) company “Deep Mind” team: AI can solve the problem of matrix multiplication. This is the first AI system to discover novel, efficient, and correct algorithms for basic tasks like matrix multiplication. In other words, the AI, called “AlphaTensor,” can discover new algorithms on its own, solving a 50-year-old problem in mathematics — finding the fastest way to multiply two matrices. Experienced stainless steel work from China
Mathematics occurs frequently in computer programming, often as a means of describing and manipulating representations of real-world phenomena. For example, it is used to represent pixels on a computer screen, weather conditions, or nodes in artificial networks. In this case, one of the main ways to use mathematics is to perform calculations on matrices.
When programming the game, the matrix describes the possible movement options. To achieve such movements, matrices are often multiplied and/or added. This takes a lot of work, especially as matrices get larger and larger, which is why computer scientists spend a lot of time and effort developing more and more efficient algorithms to get the job done. In 1969, mathematician Volker Strassen came up with a way to multiply two 2×2 matrices using only 7 multiplications instead of the standard 8. Experienced stainless steel work from China
But in the new work, “deep mind” researchers look to game systems for inspiration, most of which are based on reinforcement learning. After building some preliminary systems, the team turned their focus to tree search, which is also used in game programming as a means for the system to look at various scenarios in a given situation. When applied to multiplicative matrices, the researchers found that turning the AI system into a game searches for the most efficient way to get the desired result—a mathematical result.
Researchers test their system by allowing it to search, evaluate, and then use existing algorithms, with rewards to incentivize the selection of the most effective algorithms. The system learns the factors that affect the efficiency of matrix multiplication. Next, the researchers will allow the system to create its own algorithms, seeking to further improve efficiency. They found that, in many cases, the algorithms the system chose were better than those created by their human predecessors.
The DeepMind team hopes that in the future AI will be used more to help solve some of the most important problems in mathematics and science. Experienced stainless steel work from China