• 전기전자학회논문지
• /
• 제25권4호
• /
• pp.644-649
• /
• 2021

Self-Imitation Learning은 간단한 비활성 정책 actor-critic 알고리즘으로써 에이전트가 과거의 좋은 경험을 활용하여 최적의 정책을 찾을 수 있도록 해준다. 그리고 actor-critic 구조를 갖는 강화학습 알고리즘에 결합되어 다양한 환경들에서 알고리즘의 상당한 개선을 보여주었다. 하지만 Self-Imitation Learning이 강화학습에 큰 도움을 준다고 하더라도 그 적용 분야는 actor-critic architecture를 가지는 강화학습 알고리즘으로 제한되어 있다. 본 논문에서 Self-Imitation Learning의 알고리즘을 가치 기반 강화학습 알고리즘인 DQN에 적용하는 방법을 제안하고, Self-Imitation Learning이 적용된 DQN 알고리즘의 학습을 다양한 환경에서 진행한다. 아울러 그 결과를 기존의 결과와 비교함으로써 Self-Imitation Leaning이 DQN에도 적용될 수 있으며 DQN의 성능을 개선할 수 있음을 보인다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권3호
• /
• pp.853-873
• /
• 2021

In recent years, deep reinforcement learning (DRL) models are enjoying great interest as their success in a variety of challenging tasks. Deep Q-Network (DQN) is a widely used deep reinforcement learning model, which trains an intelligent agent that executes optimal actions while interacting with an environment. This model is well known for its ability to surpass skilled human players across many Atari 2600 games. Although DQN has achieved excellent performance in practice, there lacks a clear understanding of why the model works. In this paper, we present a visual analytics system for understanding deep Q-network in a non-blind matter. Based on the stored data generated from the training and testing process, four coordinated views are designed to expose the internal execution mechanism of DQN from different perspectives. We report the system performance and demonstrate its effectiveness through two case studies. By using our system, users can learn the relationship between states and Q-values, the function of convolutional layers, the strategies learned by DQN and the rationality of decisions made by the agent.

• 로봇학회논문지
• /
• 제14권3호
• /
• pp.157-162
• /
• 2019

The video game Tetris is one of most popular game and it is well known that its game rule can be modelled as MDP (Markov Decision Process). This paper presents a DQN (Deep Q-Network) based game agent for Tetris game. To this end, the state is defined as the captured image of the Tetris game board and the reward is designed as a function of cleared lines by the game agent. The action is defined as left, right, rotate, drop, and their finite number of combinations. In addition to this, PER (Prioritized Experience Replay) is employed in order to enhance learning performance. To train the network more than 500000 episodes are used. The game agent employs the trained network to make a decision. The performance of the developed algorithm is validated via not only simulation but also real Tetris robot agent which is made of a camera, two Arduinos, 4 servo motors, and artificial fingers by 3D printing.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권1호
• /
• pp.334-349
• /
• 2022

The effectiveness of Wi-Fi networks is greatly influenced by the optimization of contention window (CW) parameters. Unfortunately, the conventional approach employed by IEEE 802.11 wireless networks is not scalable enough to sustain consistent performance for the increasing number of stations. Yet, it is still the default when accessing channels for single-users of 802.11 transmissions. Recently, there has been a spike in attempts to enhance network performance using a machine learning (ML) technique known as reinforcement learning (RL). Its advantage is interacting with the surrounding environment and making decisions based on its own experience. Deep RL (DRL) uses deep neural networks (DNN) to deal with more complex environments (such as continuous state spaces or actions spaces) and to get optimum rewards. As a result, we present a new approach of CW control mechanism, which is termed as contention window threshold (CW_Threshold). It uses the DRL principle to define the threshold value and learn optimal settings under various network scenarios. We demonstrate our proposed method, known as a smart exponential-threshold-linear backoff algorithm with a deep Q-learning network (SETL-DQN). The simulation results show that our proposed SETL-DQN algorithm can effectively improve the throughput and reduce the collision rates.

• 사물인터넷융복합논문지
• /
• 제8권4호
• /
• pp.1-7
• /
• 2022

IoT 네트워크에서 클러스터와 싱크 노드 사이의 게이트웨이 역할을 하는 클러스터 헤드의 전력 관리는 IoT 단말의 수가 증가함에 따라 점점 더 중요해지고 있다. 특히 클러스터 헤드가 이동성을 가진 무선 단말인 경우, IoT 네트워크의 수명을 위하여 전력 소모를 최소화할 필요가 있다. 또한 IoT 네트워크에서의 전송 딜레이는 IoT 네트워크에서의 빠른 정보 수집을 위한 주요한 척도 중 하나이다. 본 논문에서는 IoT 네트워크에서 정보의 전송 딜레이를 고려한 저전력 버퍼 관리 기법을 제안한다. 제안하는 기법에서는 심층 강화학습 방법에서 사용되는 심층 Q 학습(Deep Q learning)를 사용하여 수신된 패킷을 포워딩하거나 폐기함으로써 전송 딜레이를 줄이면서도 소비 전력을 절약할 수 있다. 제안한 알고리즘은 비교에 사용된 기존 버퍼 관리 기법과 비교하여 Slotted ALOHA 프로토콜 기준 소모 전력 및 딜레이를 개선함을 보였다.

• KNOM Review
• /
• 제23권2호
• /
• pp.1-10
• /
• 2020

5G 네트워크의 핵심 기술 중 하나인 네트워크 기능 가상화 (NFV, Network Function Virtualization)는 유연하고 민첩한 네트워크 구축 및 운용을 가능하게 만드는 장점이 있다. 하지만, 한편으로는 수 많은 가상 자원을 생성하기 때문에 네트워크 관리를 복잡하게 만드는 원인이 된다. 일반적으로, NFV 환경에서는 가상 네트워크 기능(VNF, Virtual Network Function)들로 구성된 서비스 펑션 체이닝 (SFC, Service Function Chaining)을 통해 일련의 네트워크 기능들을 트래픽에 적용한다. 따라서 서비스 요구사항을 만족시킬 수 있도록 동적으로 SFC에 알맞은 양의 컴퓨팅 자원 또는 인스턴스를 할당하는 것이 필요하다. 본 논문에서는 SFC에서 적절한 수의 VNF 인스턴스를 운용하기 위해 강화학습 알고리즘의 하나인 Deep Q-Networks (DQN)을 이용한 Auto-scaling 방법을 제안한다. 제안하는 방법은 SFC로 유입되는 트래픽의 증감에 따라 SFC를 구성하는 다계층 (Multi-tier) 구조에서 스케일링(Scaling)이 필요한 계층을 선택하고, 스케일링을 통해 효과적으로 VNF 인스턴스들 개수를 조절한다.

• 스마트미디어저널
• /
• 제10권2호
• /
• pp.48-54
• /
• 2021

Sepsis is one of the leading causes of mortality globally, and it costs billions of dollars annually. However, treating septic patients is currently highly challenging, and more research is needed into a general treatment method for sepsis. Therefore, in this work, we propose a reinforcement learning method for learning the optimal treatment strategies for septic patients. We model the patient physiological time series data as the input for a deep recurrent Q-network that learns reliable treatment policies. We evaluate our model using an off-policy evaluation method, and the experimental results indicate that it outperforms the physicians' policy, reducing patient mortality up to 3.04%. Thus, our model can be used as a tool to reduce patient mortality by supporting clinicians in making dynamic decisions.

• 한국공간구조학회논문집
• /
• 제21권1호
• /
• pp.79-86
• /
• 2021

Control performance of a smart tuned mass damper (TMD) mainly depends on control algorithms. A lot of control strategies have been proposed for semi-active control devices. Recently, machine learning begins to be applied to development of vibration control algorithm. In this study, a reinforcement learning among machine learning techniques was employed to develop a semi-active control algorithm for a smart TMD. The smart TMD was composed of magnetorheological damper in this study. For this purpose, an 11-story building structure with a smart TMD was selected to construct a reinforcement learning environment. A time history analysis of the example structure subject to earthquake excitation was conducted in the reinforcement learning procedure. Deep Q-network (DQN) among various reinforcement learning algorithms was used to make a learning agent. The command voltage sent to the MR damper is determined by the action produced by the DQN. Parametric studies on hyper-parameters of DQN were performed by numerical simulations. After appropriate training iteration of the DQN model with proper hyper-parameters, the DQN model for control of seismic responses of the example structure with smart TMD was developed. The developed DQN model can effectively control smart TMD to reduce seismic responses of the example structure.

• 한국멀티미디어학회논문지
• /
• 제21권2호
• /
• pp.188-198
• /
• 2018

With recent development of high-speed wide-area wireless networks and wide spread of highperformance wireless devices, the demand on seamless video streaming services in Long Term Evolution (LTE) network environments is ever increasing. To meet the demand and provide enhanced Quality of Experience (QoE) with mobile users, the Dynamic Adaptive Streaming over HTTP (DASH) has been actively studied to achieve QoE enhanced video streaming service in dynamic network environments. However, the existing DASH algorithm to select the quality of requesting video segments is based on a procedural algorithm so that it reveals a limitation to adapt its performance to dynamic network situations. To overcome this limitation this paper proposes a novel quality selection mechanism based on a Deep Q-Network (DQN) model, the DQN-based DASH ABR($DQN_{ABR}$) mechanism. The $DQN_{ABR}$ mechanism replaces the existing DASH ABR algorithm with an intelligent deep learning model which optimizes service quality to mobile users through reinforcement learning. Compared to the existing approaches, the experimental analysis shows that the proposed solution outperforms in terms of adapting to dynamic wireless network situations and improving QoE experience of end users.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2018년도 추계학술발표대회
• /
• pp.715-718
• /
• 2018

최근 다층의 인공신경망 모델이 수많은 분야에 대한 해결 방안으로 제시되고 있으며 2015년 Mnih이 고안한 DQN(Deep Q Network)는 Atari game에서 인간 수준의 성능을 보여주며 많은 이들에게 놀라움을 자아냈다. 본 논문에서는 Atari DQN Model을 실내 자율주행 모바일 로봇에 적용하여 신경망 모델이 최단 경로를 추종하며 장애물 회피를 위한 행동을 학습시키기 위해 로봇이 가지는 상태 정보들을 84*84 Mat로 가공하였고 15가지의 행동을 정의하였다. 또한 Virtual world에서 신경망 모델이 실제와 유사한 현재 상태를 입력받아 가장 최적의 정책을 학습하고 Real World에 적용하는 방법을 연구하였다.