Deep reinforcement learning has shown remarkable capabilities in visual tasks, but it does not have a good generalization ability in the context of interference signals in the input images; this approach is therefore hard to be applied to trained agents in a new environment. To enable agents to distinguish between noise signals and important pixels in images, data augmentation techniques and the establishment of auxiliary networks are proven effective solutions. We introduce a novel algorithm, namely, saliency-extracted Q-value by augmentation (SEQA), which encourages the agent to explore unknown states more comprehensively and focus its attention on important information. Specifically, SEQA masks out interfering features and extracts salient features and then updates the mask decoder network with critic losses to encourage the agent to focus on important features and make correct decisions. We evaluate our algorithm on the DeepMind Control generalization benchmark (DMControl-GB), and the experimental results show that our algorithm greatly improves training efficiency and stability. Meanwhile, our algorithm is superior to state-of-the-art reinforcement learning methods in terms of sample efficiency and generalization in most DMControl-GB tasks.

Abstract Data is a kind of important asset in the digital economy and is driving the rise of data markets. Meanwhile, data markets promote data trading efficiently and improve the utilization of data. However, several challenges about data trading need to be addressed. Here, we resolve these challenges via our blockchain-based fair and decentralized data trading model. Disputes about data correctness is settled by the decentralized arbitration mechanism in our model. To ensure the fairness of data trading, we integrate a sale contract and a deterministic public-key encryption algorithm. The decentralization feature of blockchain cuts off the single-point failure for the data trading platform. In addition, we prove that the proposed protocol achieves the desirable security properties that a secure data trading protocol should have. Moreover, utilizing the smart contract in Solidity and program in Java, we implement our model and then evaluate its performance.

This paper investigate the distributed consensus problem for multi-agent systems with directed switching topologies that are subject to communication delays and finite data rate. By adopting a channel-based coding strategy with finite-level uniform quantizers, an effective distributed protocol is proposed, then using state augmentation technique, an enlarged closed-loop system without delays is obtained. Furthermore, the convergence analysis of the enlarged closed-loop system is conducted based on a novel input-to-output stability result. It is shown that, provided the duration of link failure in the directed switching network and the communication delays are bounded, then at each time instant, each agent (may be nonreciprocally) sends 5-level quantization information to each of its intimate neighbors, together with 3-level quantization information to itself suffices for ensuring consensus asymptotically.

The paper concerns with the problem of asynchronous gossip-based dynamic averaging estimation with communication constraints, where each agent estimates the local time-varying parameters individually, then random pairs of connected agents iteratively and locally perform a pairwise average of their estimations through quantized information communication. How quantization affects the evolution of the gossip-based dynamic averaging estimation algorithm is investigated. We prove that the agents' states converge to a random variable that deviates from the average of the estimated parameters. We derive a strong result about the upper bound for the asymptotic mean square error of the states, which just captures effect of the quantized precision and is independent of the network parameters.

UNLABELLED: A software package, IndexToolkit, aimed at overcoming the disadvantage of FASTA-format databases for frequent searching, is developed to utilize an indexing strategy to substantially accelerate sequence queries. IndexToolkit includes user-friendly tools and an Application Programming Interface (API) to facilitate indexing, storage and retrieval of protein sequence databases. As open source, it provides a sequence-retrieval developing framework, which is easily extensible for high-speed-request proteomic applications, such as database searching or modification discovering. We applied IndexToolkit to database searching engine pFind to demonstrate its effect. Experimental studies show that IndexToolkit is able to support significantly faster searches of protein database. AVAILABILITY: The IndexToolkit is free to use under the open source GNU GPL license. The source code and the compiled binary can be freely accessed through the website http://pfind.jdl.ac.cn/IndexToolkit. In this website, the more detailed information including screenshots and documentations for users and developers is also available.

Summary: Research in proteomics requires powerful database-searching software to automatically identify protein sequences in a complex protein mixture via tandem mass spectrometry. In this paper, we describe a novel database-searching software system called pFind (peptide/protein Finder), which employs an effective peptide-scoring algorithm that we reported earlier. The pFind server is implemented with the C++ STL, .Net and XML technologies. As a result, high speed and good usability of the software are achieved. Availability: The pFind web server can be freely accessed through the website http://pfind.jdl.ac.cn. In this website, the compiled binary of the local version is also available. Contact: [email protected]

Exponential periodicity of continuous-time neural networks with delays is investigated. Without assuming the boundedness and differentiability of the activation functions, some new sufficient conditions ensuring existence and uniqueness of periodic solution for a general class of neural systems are obtained. Discrete-time analogue of the continuous-time system with periodic input is formulated and we study their dynamical characteristics. The exponential periodicity of the continuous-time system is preserved by the discrete-time analogue without any restriction imposed on the uniform discretization step-size.

MOTIVATION: The correlation among fragment ions in a tandem mass spectrum is crucial in reducing stochastic mismatches for peptide identification by database searching. Until now, an efficient scoring algorithm that considers the correlative information in a tunable and comprehensive manner has been lacking. RESULTS: This paper provides a promising approach to utilizing the correlative information for improving the peptide identification accuracy. The kernel trick, rooted in the statistical learning theory, is exploited to address this issue with low computational effort. The common scoring method, the tandem mass spectral dot product (SDP), is extended to the kernel SDP (KSDP). Experiments on a dataset reported previously demonstrate the effectiveness of the KSDP. The implementation on consecutive fragments shows a decrease of 10% in the error rate compared with the SDP. Our software tool, pFind, using a simple scoring function based on the KSDP, outperforms two SDP-based software tools, SEQUEST and Sonar MS/MS, in terms of identification accuracy. SUPPLEMENTARY INFORMATION: http://www.jdl.ac.cn/user/yfu/pfind/index.html