Multivariate time series (MTS) forecasting plays an important role in the automation and optimization of intelligent applications. It is a challenging task, as we need to consider both complex intra-variable dependencies inter-variable dependencies. Existing works only learn temporal patterns with help single However, there are multi-scale many real-world MTS. Single make model prefer one type prominent shared patterns. In this article, propose adaptive graph neural network (MAGNN) address...

This paper proposes an improved TD3 (Twin Delayed Deep Deterministic Policy Gradient) algorithm to address the flaws of low success rate and slow training speed, when using original in mobile robot path planning dynamic environment. Firstly, prioritized experience replay transfer learning are introduced enhance efficiency, where probability beneficial experiences being sampled pool is increased, pre-trained model applied obstacle-free environment as initial for a Secondly, delay update...

Time-series forecasting is a key component in the automation and optimization of intelligent applications. It not trivial task, as there are various short-term and/or long-term temporal dependencies. Multiscale modeling has been considered promising strategy to solve this problem. However, existing multiscale models either apply an implicit way model dependencies or ignore interrelationships between subseries. In article, we propose interactive recurrent network (MiRNN) jointly capture...

Multivariate time series (MTS) forecasting has attracted much attention in many intelligent applications. It is not a trivial task, as we need to consider both intra-variable dependencies and inter-variable dependencies. However, existing works are designed for specific scenarios, require domain knowledge expert efforts, which difficult transfer between different scenarios. In this paper, propose scale-aware neural architecture search framework MTS (SNAS4MTF). A multi-scale decomposition...

this paper has put forward a new method to improve the performance of text categorization. The combines HMM (Hidden Markov Model) and SVM (Support Vector Machines). HMMs are used as feature extractor then vector is normalized input SVMs, so trained SVMs can classify unknown texts successfully. experimental results prove that more effective high classification accuracy.

Document clustering is one of the key problems in text mining and information retrieval area. It groups documents a way that maximizes similarity within clusters minimizes between different clusters. Most partitioning based algorithms are sensitive to initial centroids, result greatly depends on centroids. This paper first uses unsupervised feature selection method reduce dimension document space then proposes novel algorithm which select cluster centriods process by size density datasets....

The technology of real-time reliable multicast over a best-effort service network has become more popular recently. In this paper, new technique is introduced that integrates word interleaving, forward error correction (FEC) and automatic repeat request (ARQ) to mitigate the loss effects encountered in wired wireless Internet applications. For video sessions spanning tens routers, integration FEC ARQ as well fine tuning various parameters each mechanism will play major role move towards...

In many applications, it makes sense to solve the least square problems with nonnegative constraints. this article, we present a new multiplicative iteration that monotonically decreases value of quadratic programming (NNQP) objective function. This algorithm has simple closed form and is easily implemented on parallel machine. We prove global convergence apply solving image super-resolution color labelling problems. The experimental results demonstrate effectiveness broad applicability algorithm.

Summary In the past few years, executing high‐concurrency queries with interactive SQL query engines on Hadoop has become an important activity for many organizations. However, these systems do not adopt Multi‐Query Optimization (MQO) to accelerate process. There are two major concerns. Firstly, traditional MQO researches assume that multiple have high similarity. usually serve a variety of applications. Although from same application similarity, different applications may low so using will...

Association Rule Mining is an important data mining technique and Maximal frequent item sets essential step in the process of rule. Here presented BM-MFI, a new algorithm based on matrix, for maximal sets. Its basic idea transforming event database into matrix by operating rows columns to compress database. Using Itemset-Tidset pair can mine compressed with convenience effectiveness, therefore prevent conditional FP-tree candidate patterns. Experimental result verifies efficiency BM-MFI.

An optimized decision tree algorithm based on rough sets model is proposed. Firstly the most popular algorithms, which are model, usually partition examples too detailed to avoid negative impact caused by a few special because of classification accuracy. The inhibitory factors put forward in forming process cut branches for tree, avoiding redundant steps cutting later. Secondly condition attribute and matched every division unnecessary calculation improve efficiency algorithm.

This paper proposes an improved TD3 (Twin Delayed Deep Deterministic Policy Gradient) algorithm to address the flaws of low success rate and slow training speed in original for mobile robot path planning dynamic environments. Prioritized experience replay transfer learning are introduced reduce time. Then delay update strategy is devised OU noise added optimization exploring process, which improving planning. The tested by simulation where Turtlebot3 model as a object, ROS melodic operating...

Multivariate time series (MTS) forecasting has attracted much attention in many intelligent applications. It is not a trivial task, as we need to consider both intra-variable dependencies and inter-variable dependencies. However, existing works are designed for specific scenarios, require domain knowledge expert efforts, which difficult transfer between different scenarios. In this paper, propose scale-aware neural architecture search framework MTS (SNAS4MTF). A multi-scale decomposition...

Summary Impala system is an open source, analytic MPP database for Apache Hadoop. uses a query execution scheduling scheme that assigns near‐equal bytes retrieval tasks different hosts to ensure load balance. However, such “load balance” cannot guarantee short response time system, when there are original loads in the system. Traditional methods require either some assumptions or particular architecture, which be directly used In this paper, we present If fetches data from single table,...

An optimized multi-class classification algorithm based on SVM decision tree (SVMDT) is proposed. But by SVMDT, the generalization ability depends structure. In this paper, relativity separability measure between classes defined distribution of training samples to improve SVMDT. extended non-linear using kernel functions and experiments prove more effective feasible for accuracy.