CN112926628A - Action value determination method, device, learning framework, medium and equipment - Google Patents
Action value determination method, device, learning framework, medium and equipment Download PDFInfo
- Publication number
- CN112926628A CN112926628A CN202110127259.2A CN202110127259A CN112926628A CN 112926628 A CN112926628 A CN 112926628A CN 202110127259 A CN202110127259 A CN 202110127259A CN 112926628 A CN112926628 A CN 112926628A
- Authority
- CN
- China
- Prior art keywords
- value
- reinforcement learning
- learning model
- deep reinforcement
- action
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000009471 action Effects 0.000 title claims abstract description 164
- 238000000034 method Methods 0.000 title claims abstract description 69
- 230000006870 function Effects 0.000 claims abstract description 172
- 230000002787 reinforcement Effects 0.000 claims abstract description 133
- 230000008901 benefit Effects 0.000 claims abstract description 106
- 238000005070 sampling Methods 0.000 claims abstract description 83
- 230000003993 interaction Effects 0.000 claims abstract description 49
- 238000012545 processing Methods 0.000 claims description 33
- 230000008569 process Effects 0.000 claims description 24
- 238000012549 training Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 16
- 238000011156 evaluation Methods 0.000 abstract description 8
- 230000000875 corresponding effect Effects 0.000 description 95
- 230000007613 environmental effect Effects 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000002452 interceptive effect Effects 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 6
- 238000005457 optimization Methods 0.000 description 5
- 238000013473 artificial intelligence Methods 0.000 description 4
- 241000282414 Homo sapiens Species 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 239000003795 chemical substances by application Substances 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000008447 perception Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Human Computer Interaction (AREA)
- Medical Informatics (AREA)
- Image Analysis (AREA)
Abstract
The present disclosure relates to a method, an apparatus, a learning framework, a medium, and a device for determining an action value, the method including: acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the environment and a decision action corresponding to the environment state; aiming at each sampling data, determining an advantage function of the depth reinforcement learning model and an advantage function value corresponding to an environment state in the sampling data, and an advantage expectation of the advantage function value under a decision strategy corresponding to the sampling data, wherein probability distribution corresponding to the decision strategy is constructed based on strategy entropy parameters of the advantage function and the depth reinforcement learning model; and aiming at each sampling data, determining the action value corresponding to the sampling data according to the sampling data, the advantage function value corresponding to the sampling data, the advantage expectation and the state value function of the deep reinforcement learning model, and improving the evaluation accuracy of the decision strategy.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, a learning framework, a medium, and a device for determining an action value.
Background
With the development of random computer technology, various large models and complex machine learning models are gradually applied. The deep reinforcement learning combines the perception capability of the deep learning and the decision capability of the reinforcement learning, can be directly controlled according to the input image, and is closer to the thinking mode of human beings. In the training process of the deep reinforcement learning model, generally, a decision-making action strategy in a certain state needs to be evaluated based on an action value function, so as to facilitate strategy improvement of the deep reinforcement learning model.
In the related art, when the action value is determined, the action value is generally determined according to the advantage function value and the state function value, and in the calculation process, the expectation of the action value is estimated based on the average value of the advantage function value, so that errors are inevitably introduced, and the accuracy of the action value seriously affects the accuracy of a decision action determined by the deep reinforcement learning model.
Disclosure of Invention
This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.
In a first aspect, the present disclosure provides a method for determining an action value of a deep reinforcement learning model, the method comprising:
acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
for each sampling data, determining an advantage function value of the deep reinforcement learning model corresponding to an environment state in the sampling data and an advantage expectation of the advantage function value under a decision strategy corresponding to the sampling data, wherein a probability distribution corresponding to the decision strategy is constructed based on the advantage function and a strategy entropy parameter of the deep reinforcement learning model;
and aiming at each sampling data, determining an action value corresponding to the sampling data according to the sampling data, the advantage function value corresponding to the sampling data, the advantage expectation and the state value function of the deep reinforcement learning model.
In a second aspect, the present disclosure provides an apparatus for determining an action value of a deep reinforcement learning model, the apparatus comprising:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
a first determining module, configured to determine, for each sample data, an advantage function value corresponding to an advantage function of the deep reinforcement learning model and an environment state in the sample data, and an advantage expectation of the advantage function value under a decision policy corresponding to the sample data, where a probability distribution corresponding to the decision policy is constructed based on the advantage function and a policy entropy parameter of the deep reinforcement learning model;
and a second determining module, configured to determine, for each piece of the sample data, an action value corresponding to the sample data according to the sample data, an advantage function value corresponding to the sample data, the advantage expectation, and a state value function of the deep reinforcement learning model.
In a third aspect, the present disclosure provides a deep reinforcement learning framework, in a process of training the deep reinforcement learning framework, an action value corresponding to sampling data used for training is determined by the method for determining an action value of the deep reinforcement learning model according to the first aspect.
In a fourth aspect, the present disclosure provides a computer readable medium having stored thereon a computer program which, when executed by a processing apparatus, performs the steps of the method of the first aspect.
In a fifth aspect, the present disclosure provides an electronic device comprising:
a storage device having a computer program stored thereon;
processing means for executing the computer program in the storage means to carry out the steps of the method of the first aspect.
Therefore, in the technical scheme, probability distribution corresponding to the decision strategy is constructed based on the advantage function and the strategy entropy parameters of the depth reinforcement learning model, so that the probability distribution of the decision strategy can be represented through the advantage function, the advantage expectation of the advantage function value under the decision strategy corresponding to the sampling data can be calculated based on the advantage function, and accurate advantage expectation can be obtained. Furthermore, the action value of accurately evaluating the strategy of selecting decision action in the state can be determined based on the advantage expectation, effective updating of the strategy of the deep reinforcement learning model is realized, the efficiency of the deep reinforcement learning model is improved, meanwhile, the data calculation amount required by optimizing the deep reinforcement learning model can be reduced to a certain extent, the need for a large amount of sample data is avoided, the training difficulty is greatly reduced, the robustness of the deep reinforcement learning model is improved, and the high requirement of the deep reinforcement learning model on equipment resources is effectively reduced.
Additional features and advantages of the disclosure will be set forth in the detailed description which follows.
Drawings
The above and other features, advantages and aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numbers refer to the same or similar elements. It should be understood that the drawings are schematic and that elements and features are not necessarily drawn to scale. In the drawings:
FIG. 1 is a flow chart of a method for determining an action value of a deep reinforcement learning model according to an embodiment of the present disclosure;
FIG. 2 is a block diagram of an apparatus for determining an action value of a deep reinforcement learning model according to an embodiment of the present disclosure;
FIG. 3 illustrates a schematic diagram of an electronic device 600 suitable for use in implementing embodiments of the present disclosure.
Detailed Description
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it is to be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the disclosure are for illustration purposes only and are not intended to limit the scope of the disclosure.
It should be understood that the various steps recited in the method embodiments of the present disclosure may be performed in a different order, and/or performed in parallel. Moreover, method embodiments may include additional steps and/or omit performing the illustrated steps. The scope of the present disclosure is not limited in this respect.
The term "include" and variations thereof as used herein are open-ended, i.e., "including but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions for other terms will be given in the following description.
It should be noted that the terms "first", "second", and the like in the present disclosure are only used for distinguishing different devices, modules or units, and are not used for limiting the order or interdependence relationship of the functions performed by the devices, modules or units.
It is noted that references to "a", "an", and "the" modifications in this disclosure are intended to be illustrative rather than limiting, and that those skilled in the art will recognize that "one or more" may be used unless the context clearly dictates otherwise.
The names of messages or information exchanged between devices in the embodiments of the present disclosure are for illustrative purposes only, and are not intended to limit the scope of the messages or information.
Fig. 1 is a flowchart illustrating a method for determining an action value of a deep reinforcement learning model according to an embodiment of the present disclosure, where the method may include:
in step 11, an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment is obtained, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state and a decision action of the virtual environment.
The deep reinforcement learning model combines the perception capability of deep learning and the decision capability of reinforcement learning, obtains a high-dimensional observation through interaction between an agent (agent) and the environment at each moment, and perceives the observation by using a deep learning method to obtain a specific state characteristic representation of the observation, wherein the sampling data is used for representing that sampling is carried out at any moment in the interaction process, and the obtained specific state representation corresponding to the perception observation; then, a value function (state value function) of each state and a value function (action value function) of a state-action pair can be evaluated based on expected returns, and a decision strategy is promoted based on the two value functions, wherein the decision strategy is used for mapping the current state into a corresponding decision action; the environment will react to this decision-making action and get the next observation. The above process is continuously cycled to obtain the optimal strategy for achieving the goal, illustratively, the goal is to maximize the accumulated return.
In a possible embodiment, the interaction sequence is obtained by sampling during the interaction of a virtual object with the virtual environment, wherein the virtual object is controlled based on the deep reinforcement learning model, the deep reinforcement learning model is used for determining each decision action performed by the virtual object, and the virtual environment is the environment where the virtual object is located.
The virtual environment may be a virtual scene environment generated by a computer, for example, the virtual environment may be a game scene, and illustratively, multimedia data for interacting with a user is rendered so that the multimedia data can be rendered and displayed as the game scene, the virtual environment provides a multimedia virtual world, and the user can control actions of virtual objects through controls on an operation interface, or directly control virtual objects operable in the virtual environment and observe objects, characters, scenery and the like in the virtual environment from the perspective of the virtual objects, and interact through the virtual objects and other virtual objects and the like in the virtual environment. As another example, the virtual environment may also include other virtual objects in the scene, and the like. The virtual object may be an avatar in a virtual environment for simulating a user, which may be a human or other animal avatar, or the like.
As an example, the virtual object may perform a decision-making action in a first state of the virtual environment, and after the virtual object performs the decision-making action, the virtual environment may react to the decision-making action to obtain a second state of the virtual environment and a return value corresponding to the decision-making action. When sampling is performed during the interaction between the virtual object and the virtual environment, the first state, the decision action, the second state, and the return value may be used as sampling data corresponding to the sampling time, and if not otherwise stated, the environmental state in the embodiment of the present disclosure is the first state. In a complete interactive process, the sampling data according to the sequence of the sampling time is formed into an interactive sequence. For example, the virtual scene may be a virtual maze scene in which virtual rewards may appear in random positions, and the deep reinforcement learning model may be trained to determine the strategy of virtual objects from virtual maze entrance E1 to exit E2 so that the virtual objects get the most virtual rewards from entrance E1 to exit E2. Illustratively, if the sampling is performed at the entry E1 at the initial time, and the corresponding action of the virtual environment at the first state of the initial time is a straight line or a right turn, the corresponding decision action at the state of the initial time can be determined according to the policy, illustratively, the decision action is a straight line, and the environment reacts to obtain the return value and the second state based on the decision action, and the sampling obtains a sampling data. Sampling at the next moment, obtaining a first state of the virtual environment of the virtual object at the next moment, where the corresponding action in the first state is a straight movement or a right turn, determining a corresponding decision action in the first state at the next moment according to a policy, where an example decision action is a right turn, and similarly, obtaining a return value and a second state based on the response of the environment based on the decision action, and obtaining next sampling data. Then an interactive sequence containing a plurality of sampled data can be obtained by sampling in the above manner during the course of the virtual object movement value exit E2.
When sampling is performed, an image of the virtual environment corresponding to the sampling time can be acquired, so that feature extraction can be performed on the image to obtain the first state. After the virtual object performs the decision-making action, an image of the virtual environment is acquired and feature extraction is performed on the image to obtain a second state. The reward value may be a change of a score value corresponding to the virtual object after the decision action is executed, or a change of a virtual life bar, and the reward value may be set according to an actual usage scenario, which is not limited by the present disclosure.
In step 12, for each sample data, an advantage function value of the deep reinforcement learning model corresponding to the environment state in the sample data and an advantage expectation of the advantage function value under a decision strategy corresponding to the sample data are determined, wherein a probability distribution corresponding to the decision strategy is constructed based on the advantage function and the strategy entropy parameters of the deep reinforcement learning model. Entropy is a measure of uncertainty, the larger the uncertainty is, the larger the entropy is, the strategy entropy parameter in the present disclosure is a hyper-parameter in the deep reinforcement learning model, and is used for representing diversity of strategies, and the larger the strategy entropy parameter is, the more diverse the selectable strategies are represented. The value of the hyper-parameter can be set based on human experience, and can also be dynamically adjusted based on a corresponding interaction sequence in the process of updating the deep reinforcement learning model, so that the accuracy and efficiency of updating the deep reinforcement model are further improved.
In the deep reinforcement learning model, the calculation of the merit function may be implemented by a Neural network, and the merit function network may be implemented based on CNN (Convolutional Neural Networks) and RNN (Recurrent Neural Networks), for example. Therefore, the environmental state in the sample data can be input into the dominance function network, so that the output value of the dominance function network, that is, the dominance function value corresponding to the environmental state can be obtained.
In the art, the merit function is typically used to evaluate the merit value of selecting action a in state s, and therefore, in determining the merit value, it is typically necessary to determine the merit expectation of selecting each action in state s. As described in the background art, if the expectation is estimated based on the average value of the merit function values, errors are inevitably introduced, and thus, calculation errors of the action values are caused, and an optimal strategy cannot be found in the training process of the deep reinforcement learning model.
For example, the virtual scene may be a virtual maze scene in which virtual rewards may appear in random positions, and the deep reinforcement learning model may be trained to determine the strategy of virtual objects from virtual maze entrance E1 to exit E2 so that the virtual objects get the most virtual rewards from entrance E1 to exit E2. In the process of training the depth-enhanced learning model based on the interactive sequence, based on the sampling data in the interactive sequence, when the virtual object is in the state s in the virtual maze scene, three actions such as going straight, turning left or turning right can be selected in the state s, if an error is introduced into the action value when the three actions are evaluated according to the scheme in the related art, the evaluation of the decision strategy for selecting the decision action will generate an error, the accuracy of strategy improvement of the depth-enhanced learning model based on the interactive sequence is reduced, under the condition, more training sample data are usually needed to train the depth-enhanced learning model, even the convergence of the depth-enhanced learning model is influenced, and the problem that the depth-enhanced learning model is not converged occurs.
In the deep reinforcement learning model, the current state is mapped to the corresponding decision action based on the decision strategy, and therefore, in the embodiment of the present disclosure, the probability distribution corresponding to the decision strategy may be constructed based on the dominance function and the strategy entropy parameters of the deep reinforcement learning model, that is, the function of the decision strategy is solved through the dominance function and the strategy entropy parameters, so that the function of the decision strategy may have an explicit expression form, so that the dominance expectation of the dominance function value under the decision strategy corresponding to the sample data may be directly calculated based on the dominance function and the corresponding decision strategy, without using mean approximation of the sample to determine the dominance expectation. Therefore, when the action value is calculated, the problem that errors are introduced in calculation of the advantage expectation is solved, so that the accurate advantage expectation and the action value are obtained, the accurate evaluation of the decision strategy for selecting the decision action is realized, and the strategy improvement efficiency of the deep reinforcement learning model is improved.
In step 13, for each sample data, an action value corresponding to the sample data is determined according to the sample data, the merit function value corresponding to the sample data, the merit expectation, and the state value function of the deep reinforcement learning model.
In the deep reinforcement learning model, a value function is usually used to evaluate the value of a certain state or a state-action, that is, the value of an agent selecting a certain state or executing a certain action in a certain state. The value of a state is usually evaluated by using a state value function, and the value of a state can be expressed by the values of all actions in the state, i.e. the expectation of the accumulated return obtained based on the state s, and under this strategy, the accumulated return obeys a distribution, and the expectation of the accumulated return at the state is defined as a state value function v(s):
Vπ(s)＝Eπ[Gt|St＝s]
i.e. representing the state S at time t under strategy pitWhen s is the value, G is accumulatedtThe expected value at s.
As an example, the accumulated reward may be a sum of reward values corresponding to each decision-making action included in the interaction sequence. As another example, the more distant a decision-making action has an effect on the current decision-making action, the less the accumulated reward is, the accumulated reward may be the accumulated sum of the return value of each decision-making action in the interaction sequence multiplied by the attenuation coefficient corresponding to the decision-making action, wherein the attenuation coefficients corresponding to the decision-making actions decrease in the order of the decision-making actions, for example:
Gt＝Rt+1+γRt+2+γ2Rt+3+…+γn-1Rt+n
＝Rt+1+γ(Rt+2+γRt+3+…+γn-2Rt+n)
＝Rt+1+γGt+1
wherein R isiA return value for the decision action at time i, γ for the attenuation factor, and n for the number of samples in the interaction sequence after time t to the end of the interaction.
Thus, in another embodiment, the accumulated reward may be obtained from the last decision-making action of the interaction sequence by multiplying its reward value by the decay value and adding the reward value of the previous decision-making action until the reward value of the first decision-making action in the interaction sequence is added. Wherein the attenuation value can be set according to the actual usage scenario.
Similarly, in the deep reinforcement learning model, the calculation of the state value function can be realized through a neural network. Therefore, the environmental state in the sample data can be input into the state value function network, so that the output value of the state value function network, that is, the state value of the state value function corresponding to the environmental state can be obtained.
In the deep reinforcement learning model, an action value function is generally adopted to evaluate the value of executing a certain action in a certain state, that is, the expectation of the accumulated return obtained after selecting an action a in a state s is based on:
Qπ(s,a)＝Eπ[Gt|St＝s,At＝a]；
i.e. representing the state S at time t under strategy pitValue s, selected action AtWhen a, the accumulated reward GtThe expected value at s, a, that is to say the action value function, can be used to evaluate the strategy pi.
Based on the definition of the dominance function a (s, a), the state value function v(s), and the action value function Q (s, a), the following relationships exist in deep reinforcement learning in the art:
Q(s,a)＝A(s,a)+V(s)
then in this embodiment, the action value corresponding to the action value function and the environmental state and the decision action may be determined according to the sampled data, the merit function value corresponding to the sampled data, the merit expectation, and the state value function of the deep reinforcement learning model.
Therefore, in the technical scheme, probability distribution corresponding to the decision strategy is constructed based on the advantage function and the strategy entropy parameters of the depth reinforcement learning model, so that the probability distribution of the decision strategy can be represented through the advantage function, the advantage expectation of the advantage function value under the decision strategy corresponding to the sampling data can be calculated based on the advantage function, and accurate advantage expectation can be obtained. Furthermore, the action value of accurately evaluating the strategy of selecting decision action in the state can be determined based on the advantage expectation, effective updating of the strategy of the deep reinforcement learning model is realized, the efficiency of the deep reinforcement learning model is improved, meanwhile, the data calculation amount required by optimizing the deep reinforcement learning model can be reduced to a certain extent, the need for a large amount of sample data is avoided, the training difficulty is greatly reduced, the robustness of the deep reinforcement learning model is improved, and the high requirement of the deep reinforcement learning model on equipment resources is effectively reduced.
In a possible embodiment, the probability distribution corresponding to the decision policy may be constructed by:
and determining a target parameter corresponding to the decision strategy according to the advantage function value and the strategy entropy parameter. As described above, the merit function value is used to represent the value of a certain state, that is, the value of all actions in the state can be used to represent, and the merit function value may be a vector, where each dimension in the vector is used to represent the value of the action corresponding to the dimension. For example, the ratio of the merit function value to the policy entropy parameter may be determined as the target parameter corresponding to the decision policy.
Then, the probability distribution obtained after the softmax processing is performed on the target parameter may be determined as the probability distribution corresponding to the decision policy.
The manner of performing softmax processing to obtain probability distribution is conventional operation in the art, and is not described herein again. In this embodiment, the target parameters are converted into probability distribution, so that probability information for each action in the policy in a state is determined based on the dominance function value, the probability distribution of the decision policy is expressed based on the dominance function value, the decision action is subsequently determined, and the training efficiency and accuracy of the deep reinforcement learning model are improved.
In one possible embodiment, in step 13, an exemplary implementation manner of determining an action value corresponding to the sample data according to the sample data, the advantage function value corresponding to the sample data, the advantage expectation, and the state value function of the deep reinforcement learning model is as follows, and the step may include:
and determining the state value corresponding to the state value function according to the environment state in the sampling data. Illustratively, the environmental state may be input into a state value function network as described above, thereby obtaining the state value.
Then, the difference between the advantage function value and the advantage expectation is determined as a process advantage function value, so that the advantage of the value of each action in the environment state relative to the advantage expectation value can be obtained through the process advantage function value. If the value of a certain action under the environment state is better than the expected value, the processing advantage function value is a positive value, which means that the action corresponding to the processing advantage function value is selected to be positive, and more returns can be obtained. Through the conversion scheme, the processing advantage function value can meet the constraint that the processing advantage function value is expected to be 0, the stability of output is improved, the learning efficiency is improved, and meanwhile, the learning process of the deep reinforcement learning model is more stable.
Determining a sum of the processing merit function value and the state value as the action value.
As shown above, Q (s, a) ═ a (s, a) + v(s), the action value can be determined after the process merit function value and the state value are determined.
Illustratively, as indicated above, the state value V may be determined based on a state value function network, namely:
V＝V(st)
determining a dominance value A based on a dominance function network, namely:
A＝A(st)
the function of the decision strategy pi can be expressed as:
π＝P(A/τ)
wherein P represents a determined probability distribution, and τ represents the policy entropy parameter;
the processing merit function value may be further determined
Thereby determining the action value Q:
by means of the technical scheme, in the process of determining the action value corresponding to the sampling data, the state value and the advantage value are obtained by respectively evaluating the state and the state-action, and the application range of the action value determination method is widened. Meanwhile, in the process, the advantage function value is processed through the advantage expectation, so that the stability of the determined processing advantage function value can be improved, the advantage expectation is determined based on mathematical calculation, other errors cannot be introduced in the determination process, the accuracy of the action value can be guaranteed, the strategy of selecting the action in the environmental state is accurately evaluated, the strategy of the deep reinforcement learning model is effectively updated, the error of strategy updating in the deep reinforcement learning model caused by the error determined by the action value is avoided, and the accuracy of the decision action of controlling the virtual object based on the deep reinforcement learning model is guaranteed. In addition, the learning efficiency can be improved, and meanwhile, the required calculation amount and the number of samples in the process of training the deep reinforcement learning model can be reduced to a certain extent.
In one possible embodiment, the method may further comprise:
determining updated gradient information of an action value function of the deep reinforcement learning model based on the action value.
The update gradient of the action value function can be determined by deriving the parameters of the deep reinforcement learning model through a loss function corresponding to the action value function. For example, when determining the loss function, a mean square error between the target value corresponding to the action value and the action value may be calculated, such as:
Q(θ)＝Eπ[(Qπ(st,at)-Qθ(st,at))2]
wherein Q (θ) is used to represent the evaluation loss, Qθ(st,at) For expressing action value, Qπ(st,at) For representation and Qθ(st,at) And the corresponding target value theta is used for representing the model parameters to be updated in the deep reinforcement learning model.
Then, by performing derivation on the loss function and simplifying the process, such as simplifying a constant multiple formed by the derivation, the update gradient of the action value function is obtained as follows:
thus, the corresponding update gradient information can be determined based on the above formula and the environmental state and decision action in the sampled data.
And then, updating the deep reinforcement learning model according to the updating gradient information.
For example, a PPO (proximity Policy Optimization) algorithm may be used to update parameters in the deep reinforcement learning model based on the updated gradient information, so as to implement Policy Optimization of the deep reinforcement learning model.
Therefore, by the technical scheme, the updating gradient information of the deep reinforcement learning model can be determined based on the accurate action value, so that the accuracy of the determined updating gradient information is ensured, and the deviation of the strategy optimization direction caused by the error introduced into the value evaluation of the decision action is avoided. Based on the determined update gradient information, the deep reinforcement learning model can be ensured to be optimized in the direction of improving the strategy faster in the process of strategy optimization, so that the training efficiency of the deep reinforcement learning model can be effectively improved.
In one possible embodiment, an exemplary implementation manner of determining the updated gradient information of the action value function of the deep reinforcement learning model based on the action value is as follows, and the step may include:
and determining the updating gradient information of the action value function according to the updating gradient information of the decision strategy and the expected value of the difference value of the action value and the state value under the decision strategy in the component of the target direction.
The target direction may be an update gradient direction of the decision policy, and the determination manner of the action value and the state value is described in detail above and is not described herein again. In this embodiment, the deviation of the policy evaluation may be represented by an expected value of the difference between the action value and the state value in the target direction of the difference value under the decision policy. Thus, the difference between the updated gradient information of the decision strategy and the expected value may be used to construct updated gradient information of an action value function, such as:
wherein the content of the first and second substances,
updated gradient information for representing the action value function,
updated gradient information, Q, for representing the decision strategytFor representing the value of the action at time t, VtThe method is used for representing the state value corresponding to the time t, g is used for representing the updating gradient direction, and pi is used for representing the decision strategy.
Therefore, the updated gradient information of the action value function can be calculated and solved based on the relation, meanwhile, when the deep reinforcement learning model is updated based on the updated gradient information, the strategy can be guaranteed to be improved, and the error of strategy evaluation is reduced, so that the strategy determined by the updated deep reinforcement learning model is more optimal, the evaluation of the determined strategy is more accurate, the optimization efficiency of the deep reinforcement learning model is improved, and the convergence of the deep reinforcement learning model can be guaranteed to a certain extent.
The present disclosure also provides a deep reinforcement learning framework, wherein in the process of training the deep reinforcement learning framework, an action value corresponding to the sampling data for training is determined by the method for determining an action value of the deep reinforcement learning model described above. By way of example, the game artificial intelligence can be trained based on the deep reinforcement learning framework, and then through the technical scheme, the accuracy of the determined game artificial intelligence decision can be guaranteed, the decision-making capability of the game artificial intelligence is improved when the game artificial intelligence interacts with a user, and the user interaction experience is improved.
The present disclosure also provides an apparatus for determining an action value of a deep reinforcement learning model, as shown in fig. 2, the apparatus 10 includes:
an obtaining module 100, configured to obtain an interaction sequence generated by interaction between a deep reinforcement learning model and a virtual environment, where the interaction sequence includes a plurality of sample data, where each sample data includes an environment state of the virtual environment and a decision action corresponding to the environment state;
a first determining module 200, configured to determine, for each sample data, an advantage function value corresponding to an advantage function of the deep reinforcement learning model and an environmental state in the sample data, and an advantage expectation of the advantage function value under a decision policy corresponding to the sample data, where a probability distribution corresponding to the decision policy is constructed based on the advantage function and a policy entropy parameter of the deep reinforcement learning model;
a second determining module 300, configured to determine, for each of the sample data, an action value corresponding to the sample data according to the sample data, an advantage function value corresponding to the sample data, the advantage expectation, and a state value function of the deep reinforcement learning model.
Optionally, the probability distribution corresponding to the decision policy is constructed by:
determining a target parameter corresponding to the decision strategy according to the advantage function value and the strategy entropy parameter;
and determining probability distribution obtained after the target parameter is subjected to softmax processing as probability distribution corresponding to the decision strategy.
Optionally, the second determining module includes:
the first determining submodule is used for determining a state value corresponding to the state value function according to the environment state in the sampling data;
a second determining sub-module for determining a difference between the merit function value and the merit expectation as a processing merit function value;
a third determining submodule configured to determine a sum of the processing merit function value and the state value as the action value.
Optionally, the apparatus further comprises:
a third determination module, configured to determine update gradient information of an action value function of the deep reinforcement learning model based on the action value;
and the updating module is used for updating the deep reinforcement learning model according to the updating gradient information.
Optionally, the third determining module is configured to determine update gradient information of the action value function according to the update gradient information of the decision policy and an expected value of a difference between the action value and the state value in the target direction under the decision policy.
Optionally, the interaction sequence is obtained by sampling in a process of interacting a virtual object with the virtual environment, where the virtual object is controlled based on the deep reinforcement learning model, and the virtual environment is an environment where the virtual object is located.
Referring now to FIG. 3, a block diagram of an electronic device 600 suitable for use in implementing embodiments of the present disclosure is shown. The terminal device in the embodiments of the present disclosure may include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP (portable multimedia player), a vehicle terminal (e.g., a car navigation terminal), and the like, and a stationary terminal such as a digital TV, a desktop computer, and the like. The electronic device shown in fig. 3 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 3, electronic device 600 may include a processing means (e.g., central processing unit, graphics processor, etc.) 601 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage means 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data necessary for the operation of the electronic apparatus 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.
Generally, the following devices may be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 608 including, for example, tape, hard disk, etc.; and a communication device 609. The communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While fig. 3 illustrates an electronic device 600 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer readable medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 609, or may be installed from the storage means 608, or may be installed from the ROM 602. The computer program, when executed by the processing device 601, performs the above-described functions defined in the methods of the embodiments of the present disclosure.
It should be noted that the computer readable medium in the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In contrast, in the present disclosure, a computer readable signal medium may comprise a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.
In some embodiments, the clients, servers may communicate using any currently known or future developed network Protocol, such as HTTP (HyperText Transfer Protocol), and may interconnect with any form or medium of digital data communication (e.g., a communications network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network.
The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.
The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state; for each sampling data, determining an advantage function value of the deep reinforcement learning model corresponding to an environment state in the sampling data and an advantage expectation of the advantage function value under a decision strategy corresponding to the sampling data, wherein a probability distribution corresponding to the decision strategy is constructed based on the advantage function and a strategy entropy parameter of the deep reinforcement learning model; and aiming at each sampling data, determining an action value corresponding to the sampling data according to the sampling data, the advantage function value corresponding to the sampling data, the advantage expectation and the state value function of the deep reinforcement learning model.
Computer program code for carrying out operations for the present disclosure may be written in any combination of one or more programming languages, including but not limited to an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules described in the embodiments of the present disclosure may be implemented by software or hardware. The name of a module does not constitute a limitation to the module itself in some cases, for example, the obtaining module may also be described as a module for obtaining an interaction sequence generated by the deep reinforcement learning model interacting with the virtual environment.
The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), systems on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Example 1 provides a method for determining an action value of a deep reinforcement learning model, according to one or more embodiments of the present disclosure, wherein the method includes:
acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
for each sampling data, determining an advantage function value of the deep reinforcement learning model corresponding to an environment state in the sampling data and an advantage expectation of the advantage function value under a decision strategy corresponding to the sampling data, wherein a probability distribution corresponding to the decision strategy is constructed based on the advantage function and a strategy entropy parameter of the deep reinforcement learning model;
and aiming at each sampling data, determining an action value corresponding to the sampling data according to the sampling data, the advantage function value corresponding to the sampling data, the advantage expectation and the state value function of the deep reinforcement learning model.
Example 2 provides the method of example 1, wherein the probability distribution corresponding to the decision policy is constructed by:
determining a target parameter corresponding to the decision strategy according to the advantage function value and the strategy entropy parameter;
and determining probability distribution obtained after the target parameter is subjected to softmax processing as probability distribution corresponding to the decision strategy.
Example 3 provides the method of example 1, wherein the determining the action value corresponding to the sample data according to the sample data, the merit function value corresponding to the sample data, the merit expectation, and the state value function of the deep reinforcement learning model comprises:
determining a state value corresponding to the state value function according to the environment state in the sampling data;
determining a difference between the merit function value and the merit expectation as a treatment merit function value;
determining a sum of the processing merit function value and the state value as the action value.
Example 4 provides the method of example 1, wherein the method further comprises:
determining updated gradient information of an action value function of the deep reinforcement learning model based on the action value;
and updating the deep reinforcement learning model according to the updating gradient information.
Example 5 provides the method of example 4, wherein the determining updated gradient information for the action value function of the deep reinforcement learning model based on the action value includes:
and determining the updating gradient information of the action value function according to the updating gradient information of the decision strategy and the expected value of the difference value of the action value and the state value under the decision strategy in the component of the target direction.
Example 6 provides the method of example 1, wherein the interaction sequence is obtained by sampling during interaction of a virtual object with the virtual environment, wherein the virtual object is controlled based on the deep reinforcement learning model, and the virtual environment is an environment in which the virtual object is located.
Example 7 provides an apparatus for determining an action value of a deep reinforcement learning model, the apparatus comprising:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
a first determining module, configured to determine, for each sample data, an advantage function value corresponding to an advantage function of the deep reinforcement learning model and an environment state in the sample data, and an advantage expectation of the advantage function value under a decision policy corresponding to the sample data, where a probability distribution corresponding to the decision policy is constructed based on the advantage function and a policy entropy parameter of the deep reinforcement learning model;
and a second determining module, configured to determine, for each piece of the sample data, an action value corresponding to the sample data according to the sample data, an advantage function value corresponding to the sample data, the advantage expectation, and a state value function of the deep reinforcement learning model.
Example 8 provides a deep reinforcement learning framework in which, during training, an action value corresponding to sample data used for training is determined by the method for determining an action value of a deep reinforcement learning model according to any one of examples 1 to 6.
Example 9 provides a computer readable medium having stored thereon a computer program that, when executed by a processing apparatus, performs the steps of the method of any of examples 1-6, in accordance with one or more embodiments of the present disclosure.
Example 10 provides, in accordance with one or more embodiments of the present disclosure, an electronic device comprising:
a storage device having a computer program stored thereon;
processing means for executing the computer program in the storage means to carry out the steps of the method of any of examples 1-6.
The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the disclosure herein is not limited to the particular combination of features described above, but also encompasses other embodiments in which any combination of the features described above or their equivalents does not depart from the spirit of the disclosure. For example, the above features and (but not limited to) the features disclosed in this disclosure having similar functions are replaced with each other to form the technical solution.
Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limitations on the scope of the disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims. With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
Claims (10)
1. A method for determining an action value of a deep reinforcement learning model, the method comprising:
acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, wherein the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
for each sampling data, determining an advantage function value of the deep reinforcement learning model corresponding to an environment state in the sampling data and an advantage expectation of the advantage function value under a decision strategy corresponding to the sampling data, wherein a probability distribution corresponding to the decision strategy is constructed based on the advantage function and a strategy entropy parameter of the deep reinforcement learning model;
and aiming at each sampling data, determining an action value corresponding to the sampling data according to the sampling data, the advantage function value corresponding to the sampling data, the advantage expectation and the state value function of the deep reinforcement learning model.
2. The method of claim 1, wherein the probability distribution corresponding to the decision strategy is constructed by:
determining a target parameter corresponding to the decision strategy according to the advantage function value and the strategy entropy parameter;
and determining probability distribution obtained after the target parameter is subjected to softmax processing as probability distribution corresponding to the decision strategy.
3. The method of claim 1, wherein determining the action value corresponding to the sample data according to the sample data, the advantage function value corresponding to the sample data, the advantage expectation, and the state value function of the deep reinforcement learning model comprises:
determining a state value corresponding to the state value function according to the environment state in the sampling data;
determining a difference between the merit function value and the merit expectation as a treatment merit function value;
determining a sum of the processing merit function value and the state value as the action value.
4. The method of claim 1, further comprising:
determining updated gradient information of an action value function of the deep reinforcement learning model based on the action value;
and updating the deep reinforcement learning model according to the updating gradient information.
5. The method of claim 4, wherein determining updated gradient information for an action value function of the deep reinforcement learning model based on the action value comprises:
and determining the updating gradient information of the action value function according to the updating gradient information of the decision strategy and the expected value of the difference value of the action value and the state value under the decision strategy in the component of the target direction.
6. The method according to claim 1, wherein the interaction sequence is obtained by sampling during interaction of a virtual object with the virtual environment, wherein the virtual object is controlled based on the deep reinforcement learning model, and the virtual environment is an environment in which the virtual object is located.
7. An apparatus for determining an action value of a deep reinforcement learning model, the apparatus comprising:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring an interaction sequence generated by interaction of a deep reinforcement learning model and a virtual environment, the interaction sequence comprises a plurality of sampling data, and each sampling data comprises an environment state of the virtual environment and a decision action corresponding to the environment state;
a first determining module, configured to determine, for each sample data, an advantage function value corresponding to an advantage function of the deep reinforcement learning model and an environment state in the sample data, and an advantage expectation of the advantage function value under a decision policy corresponding to the sample data, where a probability distribution corresponding to the decision policy is constructed based on the advantage function and a policy entropy parameter of the deep reinforcement learning model;
and a second determining module, configured to determine, for each piece of the sample data, an action value corresponding to the sample data according to the sample data, an advantage function value corresponding to the sample data, the advantage expectation, and a state value function of the deep reinforcement learning model.
8. A deep reinforcement learning framework is characterized in that in the process of training the deep reinforcement learning framework, the action value corresponding to the sampling data used for training is determined through the action value determination method of the deep reinforcement learning model of any one of claims 1-6.
9. A computer-readable medium, on which a computer program is stored, characterized in that the program, when being executed by processing means, carries out the steps of the method of any one of claims 1 to 6.
10. An electronic device, comprising:
a storage device having a computer program stored thereon;
processing means for executing the computer program in the storage means to carry out the steps of the method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110127259.2A CN112926628A (en) | 2021-01-29 | 2021-01-29 | Action value determination method, device, learning framework, medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110127259.2A CN112926628A (en) | 2021-01-29 | 2021-01-29 | Action value determination method, device, learning framework, medium and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112926628A true CN112926628A (en) | 2021-06-08 |
Family
ID=76168693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110127259.2A Pending CN112926628A (en) | 2021-01-29 | 2021-01-29 | Action value determination method, device, learning framework, medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112926628A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114917586A (en) * | 2022-06-01 | 2022-08-19 | 北京字跳网络技术有限公司 | Model training method, object control method, device, medium, and apparatus |
| CN116663417A (en) * | 2023-06-01 | 2023-08-29 | 中国标准化研究院 | Virtual geographic environment role modeling method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190286979A1 (en) * | 2018-03-14 | 2019-09-19 | Electronic Arts Inc. | Reinforcement Learning for Concurrent Actions |
| US20200160168A1 (en) * | 2018-11-16 | 2020-05-21 | Honda Motor Co., Ltd. | Cooperative multi-goal, multi-agent, multi-stage reinforcement learning |
| WO2020113228A1 (en) * | 2018-11-30 | 2020-06-04 | Google Llc | Controlling robots using entropy constraints |
| CN111766782A (en) * | 2020-06-28 | 2020-10-13 | 浙江大学 | Strategy selection method based on Actor-Critic framework in deep reinforcement learning |
-
2021
- 2021-01-29 CN CN202110127259.2A patent/CN112926628A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190286979A1 (en) * | 2018-03-14 | 2019-09-19 | Electronic Arts Inc. | Reinforcement Learning for Concurrent Actions |
| US20200160168A1 (en) * | 2018-11-16 | 2020-05-21 | Honda Motor Co., Ltd. | Cooperative multi-goal, multi-agent, multi-stage reinforcement learning |
| WO2020113228A1 (en) * | 2018-11-30 | 2020-06-04 | Google Llc | Controlling robots using entropy constraints |
| CN111766782A (en) * | 2020-06-28 | 2020-10-13 | 浙江大学 | Strategy selection method based on Actor-Critic framework in deep reinforcement learning |
Non-Patent Citations (1)
| Title |
|---|
| CHEN QI ETC.: ""Deep Reinforcement Learning With Discrete Normalized Advantage Functions for Resource Management in Network Slicing"", 《IEEE COMMUNICATIONS LETTERS》, vol. 23, no. 8, XP011739238, DOI: 10.1109/LCOMM.2019.2922961 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114917586A (en) * | 2022-06-01 | 2022-08-19 | 北京字跳网络技术有限公司 | Model training method, object control method, device, medium, and apparatus |
| CN116663417A (en) * | 2023-06-01 | 2023-08-29 | 中国标准化研究院 | Virtual geographic environment role modeling method |
| CN116663417B (en) * | 2023-06-01 | 2023-11-17 | 中国标准化研究院 | Virtual geographic environment role modeling method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112766497A (en) | Deep reinforcement learning model training method, device, medium and equipment | |
| CN113436620B (en) | Training method of voice recognition model, voice recognition method, device, medium and equipment | |
| CN112926628A (en) | Action value determination method, device, learning framework, medium and equipment | |
| CN112291793A (en) | Resource allocation method and device of network access equipment | |
| CN113177888A (en) | Hyper-resolution restoration network model generation method, image hyper-resolution restoration method and device | |
| CN115546293B (en) | Obstacle information fusion method and device, electronic equipment and computer readable medium | |
| CN116310582A (en) | Classification model training method, image classification method, device, medium and equipment | |
| CN114964296A (en) | Vehicle driving path planning method, device, equipment and computer readable medium | |
| CN109359727B (en) | Method, device and equipment for determining structure of neural network and readable medium | |
| CN114240506A (en) | Modeling method of multi-task model, promotion content processing method and related device | |
| CN114219078A (en) | Neural network model interactive training method and device and storage medium | |
| CN113850890A (en) | Method, device, equipment and storage medium for generating animal image | |
| CN113392018A (en) | Traffic distribution method, traffic distribution device, storage medium, and electronic device | |
| CN117241092A (en) | Video processing method and device, storage medium and electronic equipment | |
| CN113052312A (en) | Deep reinforcement learning model training method and device, medium and electronic equipment | |
| CN116306981A (en) | Policy determination method, device, medium and electronic equipment | |
| CN112949850B (en) | Super-parameter determination method, device, deep reinforcement learning framework, medium and equipment | |
| CN117351299A (en) | Image generation and model training method, device, equipment and storage medium | |
| CN113052253A (en) | Hyper-parameter determination method, device, deep reinforcement learning framework, medium and equipment | |
| CN113052252B (en) | Super-parameter determination method, device, deep reinforcement learning framework, medium and equipment | |
| CN112461239B (en) | Method, device, equipment and storage medium for planning mobile body path | |
| CN112926629B (en) | Super-parameter determination method, device, deep reinforcement learning framework, medium and equipment | |
| CN111680754B (en) | Image classification method, device, electronic equipment and computer readable storage medium | |
| CN112926735A (en) | Method, device, framework, medium and equipment for updating deep reinforcement learning model | |
| CN113177176A (en) | Feature construction method, content display method and related device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |