Achieving optimal performance in reinforcement learning requires robust policies supported by training processes that ensure both sample efficiency and stability. Modeling the policy in reproducing kernel Hilbert space (RKHS) enables efficient exploration of local optimal solutions. However, the stability of existing RKHS-based methods is hindered by significant variance in gradients, while the robustness of the learned policies is often compromised due to the sensitivity of hyperparameters. In this work, we conduct a comprehensive analysis of the significant instability in RKHS policies and reveal that the variance of the policy gradient increases substantially when a wide-bandwidth kernel is employed. To address these challenges, we propose a novel RKHS policy learning method integrated with representation learning to dynamically process observations in complex environments, enhancing the robustness of RKHS policies. Furthermore, inspired by the advantage functions, we introduce a residual layer that further stabilizes the training process by significantly reducing gradient variance in RKHS. Our novel algorithm, the Residual Kernel Policy Network (ResKPN), demonstrates state-of-the-art performance, achieving a 30% improvement in episodic rewards across complex environments.

Offline reinforcement learning faces a critical challenge in synthesizing high-reward trajectories from suboptimal datasets while robustly handling the stochasticity inherent in real-world decision-making. While combination of return-conditioned sequence models, such as Decision Transformers (DT), and dynamics programming critics shows great potential in trajectory synthesis, their deterministic action generation and scale Q value critic often fails to distinguish intentional behavioral variability from detrimental noise, leading to suboptimal policy collapse. To address this challenge, we propose the Distributional Decision Transformer (DDT), a novel framework that unifies probabilistic return distribution modeling with autoregressive action generation. DDT introduces two key innovations: (1) a Gaussian stochastic return mechanism that reparameterizes target returns as samplable distributions, enabling diverse action candidate generation; and (2) an Implicit Quantile Network (IQN) critic embedded within the deciding loop, which evaluates actions across the full spectrum of return distributions (quantiles τ ~ U(0, 1)). In D4RL benchmarks, DDT achieves state-of-the-art performance, achieving a 91.6 average normalized score in MuJoCo locomotion and 69.3 in sparse-reward settings. The results establish DDT as a principled solution for synthesis of risk-aware trajectory in offline RL.

EinStein würfelt nicht! is a perfect information stochastic game in which randomness influences the game process enormously. In this paper, we propose an optimized algorithm named Quick Neural Network Tree Search (QNNTS) based on deep reinforcement learning and Monte Carlo tree search (MCTS) to construct the artificial intelligence agent of EWN. Meanwhile the lightness of the model makes it possible to train with much less computing resources. The optimization structure of the algorithm based on MCTS is named Optimized UCT with Heuristic Search (UCTH), which introduces the expectation valuation strategy into the Monte Carlo tree search. As the prerequisite product of QNNTS, it performs with an improvement of the winning rate. Ultimately, the Attention-ResNet structure combined with domain knowledge is used to obtain the proposed algorithm. Compared with several conventional algorithms, it gains high winning rates of at least 68%.

The safe storage and sharing of medical data have promoted the development of the public medical field. At the same time, blockchain technology guarantees the safe storage and sharing of medical data. However, the consensus algorithm in the current medical blockchain cannot meet the requirements of low delay and high throughput in the large-scale network, and the identity of the primary node is exposed and vulnerable to attack. Therefore, this paper proposes an efficient consensus algorithm for medical data storage and sharing based on a master–slave multi-chain of alliance chain (ECA_MDSS). Firstly, institutional nodes in the healthcare alliance chain are clustered according to geographical location and medical system structure to form a multi-zones network. The system adopts master–slave multi-chain architecture to ensure security, and each zone processes transactions in parallel to improve consensus efficiency. Secondly, the aggregation signature is used to improve the practical Byzantine fault-tolerant (PBFT) consensus to reduce the communication interaction of consensus in each zone. Finally, an efficient ring signature is used to ensure the anonymity and privacy of the primary node in each zone and to prevent adaptive attacks. Meanwhile, a trust model is introduced to evaluate the trust degree of the node to reduce the evil done by malicious nodes. The experimental results show that ECA_ MDSS can effectively reduce communication overhead and consensus delay, improve transaction throughput, and enhance system scalability.