Search

When dealing with sparse reward problems,existing deep RL algorithms often lead to hard exploration,they often only rely on the pre-designed environment reward,so it is difficult to achieve good results.In this situation,it is necessary to design rewards more carefully,make more accurate judgments and feedback on the exploration status of agents.The asynchronous advantage actor-critic(A3C) algorithm improves the training efficiency through parallel training,and improves the training speed of the original algorithm.However,for the environment with sparse rewards,it cannot well solve the problem of difficult exploration.To solve the problem of poor exploration effect of A3C algorithm in sparse reward environment,A3C based on exploration trajectory perception(ETP-A3C) is proposed.The algorithm can perceive the exploration trajectory of the agent when it is difficult to explore in training,further judge and decide the exploration direction of the agent,and help the agent get out of the exploration dilemma as soon as possible.In order to verify the effectiveness of ETP-A3C algorithm,a comparative experiment is carried out with baseline algorithm in five different environments of Super Mario Brothers.The results show that this method has significantly improved the learning speed and model stability.

In recent years, the rapid development of artificial intelligence has enabled the launch of many new screening tools. This review aims to facilitate screening tool selection through a systematic narrative review and feature analysis. The current adoption rate of transparent tool reporting is low: by screening 191 studies published in the Review of Educational Research since 2015, we found that only eight studies reported screening tools. More research is needed to understand the reasons behind this phenomenon. After consulting various sources, 26 available screening tools in the market were found. Among them, we identified and evaluated 12 screening tools for educational reviewers and ranked them in descending order of feature score: Covidence (1), DistillerSR (2, tied), EPPI-Reviewer (2, tied), CADIMA (4), Swift-Active (5), Rayyan (6, tied), SysRev (6, tied), Abstrackr (8, tied), ReLiS (8, tied), RevMan (8, tied), ASReview (11), and Excel (12). In the discussion, we provide insights into the promise and bias in tools' machine learning algorithms. Our results encourage researchers to report their tool usage in publications and select tools based on suitability instead of convenience. [ABSTRACT FROM AUTHOR]

Ultrabroadband and polarization‐sensitive imaging are essential for pioneering advancements in intelligent technology, offering a pivotal pathway to multidimensional information extraction. However, the dearth of appropriate photosensitive semiconductors in the infrared region and the lack of suitably efficient device architecture for the separation and collection of photocarriers significantly impede the development of ultra‐broadband and polarization‐sensitive photodetectors. To address these challenges, a sandwiched Ta2NiSe5/MoTe2/Graphene heterostructure device is engineered. In this device, efficient and polarization‐sensitive photocarriers are generated in the top Ta2NiSe5 layer, rapidly separated through the middle MoTe2 layer, and effectively collected by the bottom graphene layer. As a result, the developed photodetector exhibits an ultra‐broadband and polarization‐sensitive photoresponse that extends from visible light (520 nm) to short‐wave infrared (2200 nm). At 2200 nm, the device displays a notable responsivity of 5.79 A W−1, a specific detectivity of 1010 Jones, and an anisotropic ratio of 1.44. Furthermore, this device successfully demonstrates high‐resolution ultra‐broadband and polarized light imaging capabilities. This study thus presents an intriguing blueprint for the development of advanced 2D imaging platforms for future‐generation intelligent systems. [ABSTRACT FROM AUTHOR]

Objectives: This study aims to decrease the scan time and enhance image quality in pediatric total-body PET imaging by utilizing multimodal artificial intelligence techniques. Methods: A total of 270 pediatric patients who underwent total-body PET/CT scans with a uEXPLORER at the Sun Yat-sen University Cancer Center were retrospectively enrolled. 18F-fluorodeoxyglucose (18F-FDG) was administered at a dose of 3.7 MBq/kg with an acquisition time of 600 s. Short-term scan PET images (acquired within 6, 15, 30, 60 and 150 s) were obtained by truncating the list-mode data. A three-dimensional (3D) neural network was developed with a residual network as the basic structure, fusing low-dose CT images as prior information, which were fed to the network at different scales. The short-term PET images and low-dose CT images were processed by the multimodal 3D network to generate full-length, high-dose PET images. The nonlocal means method and the same 3D network without the fused CT information were used as reference methods. The performance of the network model was evaluated by quantitative and qualitative analyses. Results: Multimodal artificial intelligence techniques can significantly improve PET image quality. When fused with prior CT information, the anatomical information of the images was enhanced, and 60 s of scan data produced images of quality comparable to that of the full-time data. Conclusion: Multimodal artificial intelligence techniques can effectively improve the quality of pediatric total-body PET/CT images acquired using ultrashort scan times. This has the potential to decrease the use of sedation, enhance guardian confidence, and reduce the probability of motion artifacts. [ABSTRACT FROM AUTHOR]

In the generation process of reinforcement learning decision model, due to the complex environment and incomplete observation of state information, the classical proximal policy optimization algorithm faces problems such as low exploration and utilization efficiency and poor effect of generated strategies, this paper proposes an MNAEK-PPO (proximal policy optimization based on maximum number of arrival & expert knowledge algorithm) based on curiosity mechanism. Focusing on the difficult problem of exploring the strategy space, by constructing the exploration frequency matrix of the agent in the training process, the exploration frequency is treated as an internal reward to participate in the agent's reinforcement learning and training process. In addition, expert knowledge is added to assist the agent in making decisions. Through experiments in the intelligent battlefield simulation environment, the best construction method of internal rewards in MNAEK-PPO is determined, and a series of comparative experiments are carried out. The experimental results show that MNAEK-PPO greatly improves the exploration efficiency of decision space, and the convergence speed and game score are significantly improved, which provides a new solution for promoting the application and development of deep reinforcement learning in the generation of intelligent tactical strategies. [ABSTRACT FROM AUTHOR]