Search

• The multi-agent defense and attack environment is reconstructed. • Several algorithms are applied to solve the considered problem. • We redefine the state space, the action space and the reward functions accordingly. • Comparison experiments are conducted to show the performance of the employed models. Learning to cooperate among agents has always been an important research topic in artificial intelligence. Multi-agent defense and attack, one of the important issues in multi-agent cooperation, requires multiple agents in the environment to learn effective strategies to achieve their goals. Deep reinforcement learning (DRL) algorithms have natural advantages dealing with continuous control problems especially under situations with dynamic interactions, and have provided new solutions for those long-studied multi-agent cooperation problems. In this paper, we start from deep deterministic policy gradient (DDPG) algorithm and then introduce multi-agent DDPG (MADDPG) to solve the multi-agent defense and attack problem under different situations. We reconstruct the considered environment, redefine the continuous state space, continuous action space, reward functions accordingly, and then apply deep reinforcement learning algorithms to obtain effective decision strategies. Several experiments considering different confrontation scenarios are conducted to validate the feasibility and effectiveness of the DRL-based methods. Experimental results show that through learning the agents can make better decisions, and learning with MADDPG achieves superior performance than learning with other DRL-based models, which also explains the importance and necessity of mastering other agents' information. [ABSTRACT FROM AUTHOR]

The core issue of learning-to-rank (LTR) for document retrieval lies in finding an optimal ranking policy to meet the search intent of the user. The majority of proposed LTR approaches treat the ranking as a static process, employing a fixed ranking policy to immediately assign scores to documents. By contrast, ranking is not a static but an interactive process where the user continues interacting with the document retrieval system through information exchange such as search intent (e.g., rating or clicking for the retrieved items). We model the interactive ranking process (IRP), and propose an Attention-Based Interactive LTR model (AIRank) to constitute an intent-aware flexible ranking policy to gratify the user’s need. To enhance the ranking quality, the inherent relations among documents are procured by the self-attention method to contribute to an enriched user intent representation. Furthermore, we mend the policy gradient learning method to train the AIRank in the IRP. Experiments demonstrate the effectiveness of AIRank compared to the state-of-the-art methods in terms of normalized discounted cumulative gain and expected reciprocal rank. [ABSTRACT FROM AUTHOR]

Although alveolar macrophages (AMs) play important roles in preventing and eliminating pulmonary infections, little is known about their regulation in healthy animals. Since exposure to LPS often renders cells hyporesponsive to subsequent LPS exposures ("tolerant"), we tested the hypothesis that LPS produced in the intestine reaches the lungs and stimulates AMs, rendering them tolerant. We found that resting AMs were more likely to be tolerant in mice lacking acyloxyacyl hydrolase (AOAH), the host lipase that degrades and inactivates LPS; isolated Aoah-/- AMs were less responsive to LPS stimulation and less phagocytic than were Aoah+/+ AMs. Upon innate stimulation in the airways, Aoah-/- mice had reduced epithelium- and macrophage-derived chemokine/cytokine production. Aoah-/- mice also developed greater and more prolonged loss of body weight and higher bacterial burdens after pulmonary challenge with Pseudomonas aeruginosa than did wildtype mice. We also found that bloodborne or intrarectally-administered LPS desensitized ("tolerized") AMs while antimicrobial drug treatment that reduced intestinal commensal Gram-negative bacterial abundance largely restored the innate responsiveness of Aoah-/- AMs. Confirming the role of LPS stimulation, the absence of TLR4 prevented Aoah-/- AM tolerance. We conclude that commensal LPSs may stimulate and desensitize (tolerize) alveolar macrophages in a TLR4-dependent manner and compromise pulmonary immunity. By inactivating LPS in the intestine, AOAH promotes antibacterial host defenses in the lung. Author summary: AOAH is the host lipase that degrades and inactivates Gram-negative bacterial lipopolysaccharides (LPSs). AOAH is required for recovery from LPS-induced macrophage tolerance. Expressed in the gut, AOAH inactivates microbiota-derived LPS. In this study we found that AOAH-deficient mice were less able to contain pulmonary Pseudomonas aeruginosa infection than were control wildtype mice. Alveolar macrophages (AMs) from Aoah-/- mice were hypo-responsive to innate stimulation and they had reduced phagocytic activity. In addition, Aoah-/- AMs had metabolic changes characteristic of tolerant macrophages as well as increased cell-surface expression of MHC II and co-stimulatory molecules, findings suggesting that they had been stimulated in situ. Treating Aoah-/- mice with p.o. neomycin normalized AMs' innate responsiveness while intrarectal LPS administration tolerized AMs. We conclude that AOAH regulates pulmonary mucosal immunity in part by inactivating LPS in the gut. This study sheds light on a previously unappreciated mechanism that regulates pulmonary immune defense via the gut-lung axis. [ABSTRACT FROM AUTHOR]