An information processing method and electronic device

Abstract

The invention discloses an information processing method and electronic equipment. The information processing method comprises the following steps of detecting the state of the electronic equipment; when detecting that the state of the electronic equipment is transferred to a third state st+n+1 from a second state st+n and determining that a second action at+n for causing the state transfer is not a do-nothing operation, removing the second state st+n from a target state set T, wherein the do-nothing operation is that a user does not operate the electronic equipment; when determining that a first action at+n-1 for causing the state transfer is not the do-nothing operation, looking forward the first do-nothing operation from the second action at+n; when determining that the third state st+n+1 appears in the state after the first searched do-nothing operation for the first time, adding the third state st+n+1 into the target state set T; and determining the state in the target state set T as a target state.

Description

technical field [0001] The present invention relates to computer technology, in particular to an information processing method and electronic equipment. Background technique [0002] In the field of enhanced learning, there is a technology called artificial intelligence reinforcement learning, such as Q-Learning technology, which allows electronic devices to learn in the process of continuous interaction with the environment, so as to find the Departs the optimal sequence of operations to reach the goal state. [0003] This technique first formalizes the problem to be solved as a Markov decision problem (Markov Decision Problem, MDP). The MDP problem can be described as: state set S, action set A, state transition model T. [0004] Each element in the state set S is an N-dimensional vector used to represent a specific state, assuming that the state of the electronic device at time t is s t =(x 1 ,x 2 ,...,x N ), s t ∈S, S is also called the state space. [0005] The ...

AI Technical Summary

Problems solved by technology

However, the target state is not always predicted in advance. When the target state is unknown, the electronic device cannot use the above technology to find the optimal operation sequence from any state to the target state.

Images