Results for “Q learning”

NASA NTRS1996

Fuzzy Q-Learning for Generalization of Reinforcement Learning

Fuzzy Q-Learning, introduced earlier by the author, is an extension of Q-Learning into fuzzy environments. GARIC is a methodology for fuzzy reinforcement learning. In this paper, we introduce GARIC-Q, a new method for doing incremental Dynamic Programming using a society of intelligent agents which are controlled at the top level by Fuzzy Q-Learning and at the local level, each agent learns and operates based on GARIC. GARIC-Q improves the speed and applicability of Fuzzy Q-Learning through generalization of input space by using fuzzy rules and bridges the gap between Q-Learning and rule based intelligent systems.

Berenji, Hamid R.↗

NASA NTRS2005

Quicker Q-Learning in Multi-Agent Systems

Multi-agent learning in Markov Decisions Problems is challenging because of the presence ot two credit assignment problems: 1) How to credit an action taken at time step t for rewards received at t' greater than t; and 2) How to credit an action taken by agent i considering the system reward is a function of the actions of all the agents. The first credit assignment problem is typically addressed with temporal difference methods such as Q-learning OK TD(lambda) The second credit assi,onment problem is typically addressed either by hand-crafting reward functions that assign proper credit to an agent, or by making certain independence assumptions about an agent's state-space and reward function. To address both credit assignment problems simultaneously, we propose the Q Updates with Immediate Counterfactual Rewards-learning (QUICR-learning) designed to improve both the convergence properties and performance of Q-learning in large multi-agent problems. Instead of assuming that an agent s value function can be made independent of other agents, this method suppresses the impact of other agents using counterfactual rewards. Results on multi-agent grid-world problems over multiple topologies show that QUICR-learning can achieve up to thirty fold improvements in performance over both conventional and local Q-learning in the largest tested systems.

Agogino, Adrian K.↗

NASA NTRSDate not supplied

Applying Machine Learning Tools for Runway Configuration Decision Support

Determining optimal runway configurations at airports, a responsibility assigned to air traffic controllers, is a challenging task. The decision-making process is intricate and involves consideration of many factors such as prevailing wind condition, convective weather, visibility, cloud ceilings, departure and arrival demand, traffic flow, equipment status, and other airport constraints. In a previous work, we developed a Runway Configuration Assistance tool using an offline reinforcement learning method called conservative Q-learning. In this paper, we evaluate and validate our Runway Configuration Assistance tool as a decision support for air traffic controllers. We validated our tool using three airports with differing levels of complexity: Charlotte Douglas International Airport, Denver International Airport, and Dallas Fort Worth International Airport. We quantified the performance of the Runway Configuration Assistance tool based on (1) agreement with historical air traffic controller decisions and (2) violation of decisions that would be obvious to subject-matter experts. Our tool showed promising results in both performance metrics for the three airports, despite the complexities in the runway configuration decision-making process. We also discuss challenges in using machine learning in general to aid air traffic management and identify deployment considerations for the Runway Configuration Assistance tool.

Runway Configuration Management↗

NASA NTRSDate not supplied

Machine Learning Tools for Runway Configuration Decision Support

Determining optimal runway configurations at airports, a responsibility assigned to air traffic controllers, is a challenging task. The decision-making process is intricate and involves consideration of many factors such as prevailing wind condition, convective weather, visibility, cloud ceilings, departure and arrival demand, traffic flow, equipment status, and other airport constraints. In a previous work, we developed a Runway Configuration Assistance tool using an offline reinforcement learning method called conservative Q-learning. In this paper, we evaluate and validate our Runway Configuration Assistance tool as a decision support for air traffic controllers. We validated our tool using three airports with differing levels of complexity: Charlotte Douglas International Airport, Denver International Airport, and Dallas Fort Worth International Airport. We quantified the performance of the Runway Configuration Assistance tool based on (1) agreement with historical air traffic controller decisions and (2) violation of decisions that would be obvious to subject-matter experts. Our tool showed promising results in both performance metrics for the three airports, despite the complexities in the runway configuration decision-making process. We also discuss challenges in using machine learning in general to aid air traffic management and identify deployment considerations for the Runway Configuration Assistance tool.

Runway Configuration Management↗

NASA NTRS2003

Reinforcement Learning for Weakly-Coupled MDPs and an Application to Planetary Rover Control

Weakly-coupled Markov decision processes can be decomposed into subprocesses that interact only through a small set of bottleneck states. We study a hierarchical reinforcement learning algorithm designed to take advantage of this particular type of decomposability. To test our algorithm, we use a decision-making problem faced by autonomous planetary rovers. In this problem, a Mars rover must decide which activities to perform and when to traverse between science sites in order to make the best use of its limited resources. In our experiments, the hierarchical algorithm performs better than Q-learning in the early stages of learning, but unlike Q-learning it converges to a suboptimal policy. This suggests that it may be advantageous to use the hierarchical algorithm when training time is limited.

Bernstein, Daniel S.↗

NASA NTRS2006

QUICR-learning for Multi-Agent Coordination

Coordinating multiple agents that need to perform a sequence of actions to maximize a system level reward requires solving two distinct credit assignment problems. First, credit must be assigned for an action taken at time step t that results in a reward at time step t > t. Second, credit must be assigned for the contribution of agent i to the overall system performance. The first credit assignment problem is typically addressed with temporal difference methods such as Q-learning. The second credit assignment problem is typically addressed by creating custom reward functions. To address both credit assignment problems simultaneously, we propose the "Q Updates with Immediate Counterfactual Rewards-learning" (QUICR-learning) designed to improve both the convergence properties and performance of Q-learning in large multi-agent problems. QUICR-learning is based on previous work on single-time-step counterfactual rewards described by the collectives framework. Results on a traffic congestion problem shows that QUICR-learning is significantly better than a Q-learner using collectives-based (single-time-step counterfactual) rewards. In addition QUICR-learning provides significant gains over conventional and local Q-learning. Additional results on a multi-agent grid-world problem show that the improvements due to QUICR-learning are not domain specific and can provide up to a ten fold increase in performance over existing methods.

Agogino, Adrian K.↗

NASA NTRSDate not supplied

Deep Reinforcement Learning based Routing in an Air-to-Air Ad-hoc Network

This paper studies the Multiple Sources and Multiple Destinations (MSMD) routing problem in a dynamic Air-to-Air Ad-hoc Network (AAAN). We consider a spectrum limited scenario where multiple links have to share the same frequency channel so that co-channel interference becomes inevitable. As a result, routing decisions and spectrum access are coupled and must be jointly considered. This paper proposes a deep Q-learning based algorithm to find an optimal routing and channel selection strategy that minimizes the end-to-end communication delay. Specifically, under the assumption that only local information is available to every node, the Deep Q-Network (DQN) is trained offline to learn the optimal routing and channel selection strategy. After the trained DQN is implemented in every node, multiple relay nodes can simultaneously determine their next-hop relay and channel selections in real-time. Simulation results demonstrate the efficacy of our proposed algorithm.

AAAN↗

NASA NTRSDate not supplied

Deep Reinforcement Learning based Routing in an Air-to-Air Ad-hoc Network

This paper studies the Multiple Sources and Multiple Destinations (MSMD) routing problem in a dynamic Air-to-Air Ad-hoc Network (AAAN). We consider a spectrum limited scenario where multiple links have to share the same frequency channel so that co-channel interference becomes inevitable. As a result, routing decisions and spectrum access are coupled and must be jointly considered. This paper proposes a deep Q-learning based algorithm to find an optimal routing and channel selection strategy that minimizes the end-to-end communication delay. Specifically, under the assumption that only local information is available to every node, the Deep Q-Network (DQN) is trained offline to learn the optimal routing and channel selection strategy. After the trained DQN is implemented in every node, multiple relay nodes can simultaneously determine their next-hop relay and channel selections in real-time. Simulation results demonstrate the efficacy of our proposed algorithm.

AAAN↗

NASA NTRS1991

On the integration of reinforcement learning and approximate reasoning for control

The author discusses the importance of strengthening the knowledge representation characteristic of reinforcement learning techniques using methods such as approximate reasoning. The ARIC (approximate reasoning-based intelligent control) architecture is an example of such a hybrid approach in which the fuzzy control rules are modified (fine-tuned) using reinforcement learning. ARIC also demonstrates that it is possible to start with an approximately correct control knowledge base and learn to refine this knowledge through further experience. On the other hand, techniques such as the TD (temporal difference) algorithm and Q-learning establish stronger theoretical foundations for their use in adaptive control and also in stability analysis of hybrid reinforcement learning and approximate reasoning-based controllers.

Berenji, Hamid R.↗

NASA NTRS2004

Unifying Temporal and Structural Credit Assignment Problems

Single-agent reinforcement learners in time-extended domains and multi-agent systems share a common dilemma known as the credit assignment problem. Multi-agent systems have the structural credit assignment problem of determining the contributions of a particular agent to a common task. Instead, time-extended single-agent systems have the temporal credit assignment problem of determining the contribution of a particular action to the quality of the full sequence of actions. Traditionally these two problems are considered different and are handled in separate ways. In this article we show how these two forms of the credit assignment problem are equivalent. In this unified frame-work, a single-agent Markov decision process can be broken down into a single-time-step multi-agent process. Furthermore we show that Monte-Carlo estimation or Q-learning (depending on whether the values of resulting actions in the episode are known at the time of learning) are equivalent to different agent utility functions in a multi-agent system. This equivalence shows how an often neglected issue in multi-agent systems is equivalent to a well-known deficiency in multi-time-step learning and lays the basis for solving time-extended multi-agent problems, where both credit assignment problems are present.

Agogino, Adrian K.↗

NASA NTRSDate not supplied

Joint Spectrum Access and Power Control in Air-Air Communications - A Deep Reinforcement Learning Based Approach

This paper considers the dynamic spectrum access and power control problem in a single-hop point-to-point Air-Air Communication Network (AACN). Due to spectrum scarcity, we assume the number of Aircraft-to-Aircraft (A2A) communication links is greater than that of the available channels, such that some communication links need to share the same channel, causing co-channel interference. We formulate the joint channel selection and power control optimization problem to maximize the Weighted Sum Spectral Efficiency (WSSE). A distributed and dynamic deep Q learning-based algorithm is proposed to find the optimal solution. Specifically, we design two different policies that are trained by conducting a trial-and-error scheme. Each communication link can achieve the optimal policy by exploiting the local information from its neighbors, and this distributive approach make it scalable to large networks. Finally, our experimental results demonstrate the effectiveness of the proposed solution in various AACN scenarios.

Zhe Wang↗

NASA NTRSDate not supplied

Joint Spectrum Access and Power Control in Air-Air Communications - A Deep Reinforcement Learning Based Approach

This paper considers the dynamic spectrum access and power control problem in a single-hop point-to-point Air-Air Communication Network (AACN). Due to spectrum scarcity, we assume the number of Aircraft-to-Aircraft (A2A) communication links is greater than that of the available channels, such that some communication links need to share the same channel, causing co-channel interference. We formulate the joint channel selection and power control optimization problem to maximize the Weighted Sum Spectral Efficiency (WSSE). A distributed and dynamic deep Q learning-based algorithm is proposed to find the optimal solution. Specifically, we design two different policies that are trained by conducting a trial-and-error scheme. Each communication link can achieve the optimal policy by exploiting the local information from its neighbors, and this distributive approach make it scalable to large networks. Finally, our experimental results demonstrate the effectiveness of the proposed solution in various AACN scenarios.

Zhe Wang↗

NASA NTRSDate not supplied

Learning Based Edge Computing in Air-to-Air Communication Network

This paper studies learning-based edge computing and communication in a dynamic Air-to-Air Ad-hoc Network (AAAN). Due to spectrum scarcity, we assume the number of Air-to-Air (A2A) communication links is greater than that of the available frequency channels, such that some communication links have to share the same channel, causing co-channel interference. We formulate the joint channel selection and power control optimization problem to maximize the aggregate spectrum utilization efficiency under resource and fairness constraints. A distributed deep Q learning-based edge computing and communication algorithm is proposed to find the optimal solution. In particular, we design two different neural network structures and each communication link can converge to the optimal operation by exploiting only the local information from its neighbors, making it scalable to large networks. Finally, experimental results demonstrate the effectiveness of the proposed solution in various AAAN scenarios.

Zhe Wang↗

NASA NTRS1997

Collaborating Fuzzy Reinforcement Learning Agents

Earlier, we introduced GARIC-Q, a new method for doing incremental Dynamic Programming using a society of intelligent agents which are controlled at the top level by Fuzzy Relearning and at the local level, each agent learns and operates based on ANTARCTIC, a technique for fuzzy reinforcement learning. In this paper, we show that it is possible for these agents to compete in order to affect the selected control policy but at the same time, they can collaborate while investigating the state space. In this model, the evaluator or the critic learns by observing all the agents behaviors but the control policy changes only based on the behavior of the winning agent also known as the super agent.

Berenji, Hamid R.↗

NASA NTRS1997

Learning and Tuning of Fuzzy Rules

In this chapter, we review some of the current techniques for learning and tuning fuzzy rules. For clarity, we refer to the process of generating rules from data as the learning problem and distinguish it from tuning an already existing set of fuzzy rules. For learning, we touch on unsupervised learning techniques such as fuzzy c-means, fuzzy decision tree systems, fuzzy genetic algorithms, and linear fuzzy rules generation methods. For tuning, we discuss Jang's ANFIS architecture, Berenji-Khedkar's GARIC architecture and its extensions in GARIC-Q. We show that the hybrid techniques capable of learning and tuning fuzzy rules, such as CART-ANFIS, RNN-FLCS, and GARIC-RB, are desirable in development of a number of future intelligent systems.

Berenji, Hamid R.↗

NASA NTRS1986

Investigation of porous polymer gas chromatographic packings for atmospheric analysis of extraterrestrial bodies

Measurement of the permanent gases in the atmospheres of mission targets is a major objective. A 16 meter long Porapak N column was used on the Venus probe and required a rather high carrier gas flow rate. The researchers have, therefore, surveyed commercial porous polymer types which had some ability to resolve nitrogen, oxygen, argon and carbon monoxide gases. Porapaks N and Q appeared superior to most. Batch to batch variation, however, was quite wide, so the researchers learned how to synthesize porous polymer and investigated some of the factors affecting the separations. A polymer was synthesized which was superior to all commercial products and allowed at least a 50% reduction in length and flow rate of carrier gas. Similar studies were made concerning the separation of hydrocarbons and new porous polymers have been synthesized which represent significant improvements in time of analysis, column, and carrier gas flow rate.

Pollock, G. E.↗

NASA NTRS1992

Using neural networks and Dyna algorithm for integrated planning, reacting and learning in systems

The traditional AI answer to the decision making problem for a robot is planning. However, planning is usually CPU-time consuming, depending on the availability and accuracy of a world model. The Dyna system generally described in earlier work, uses trial and error to learn a world model which is simultaneously used to plan reactions resulting in optimal action sequences. It is an attempt to integrate planning, reactive, and learning systems. The architecture of Dyna is presented. The different blocks are described. There are three main components of the system. The first is the world model used by the robot for internal world representation. The input of the world model is the current state and the action taken in the current state. The output is the corresponding reward and resulting state. The second module in the system is the policy. The policy observes the current state and outputs the action to be executed by the robot. At the beginning of program execution, the policy is stochastic and through learning progressively becomes deterministic. The policy decides upon an action according to the output of an evaluation function, which is the third module of the system. The evaluation function takes the following as input: the current state of the system, the action taken in that state, the resulting state, and a reward generated by the world which is proportional to the current distance from the goal state. Originally, the work proposed was as follows: (1) to implement a simple 2-D world where a 'robot' is navigating around obstacles, to learn the path to a goal, by using lookup tables; (2) to substitute the world model and Q estimate function Q by neural networks; and (3) to apply the algorithm to a more complex world where the use of a neural network would be fully justified. In this paper, the system design and achieved results will be described. First we implement the world model with a neural network and leave Q implemented as a look up table. Next, we use a lookup table for the world model and implement the Q function with a neural net. Time limitations prevented the combination of these two approaches. The final section discusses the results and gives clues for future work.

Lima, Pedro↗

NASA NTRS1999

Lesson Plan Prototype for International Space Station's Interactive Video Education Events

The outreach and education components of the International Space Station Program are creating a number of materials, programs, and activities that educate and inform various groups as to the implementation and purposes of the International Space Station. One of the strategies for disseminating this information to K-12 students involves an electronic class room using state of the art video conferencing technology. K-12 classrooms are able to visit the JSC, via an electronic field trip. Students interact with outreach personnel as they are taken on a tour of ISS mockups. Currently these events can be generally characterized as: Being limited to a one shot events, providing only one opportunity for students to view the ISS mockups; Using a "one to many" mode of communications; Using a transmissive, lecture based method of presenting information; Having student interactions limited to Q&A during the live event; Making limited use of media; and Lacking any formal, performance based, demonstration of learning on the part of students. My project involved developing interactive lessons for K-12 students (specifically 7th grade) that will reflect a 2nd generation design for electronic field trips. The goal of this design will be to create electronic field trips that will: Conform to national education standards; More fully utilize existing information resources; Integrate media into field trip presentations; Make support media accessible to both presenters and students; Challenge students to actively participate in field trip related activities; and Provide students with opportunities to demonstrate learning

Zigon, Thomas↗