With geological big data becoming a focus of geoscience research, the vast amount textual provides both opportunities and challenges for analysis mining. In fact, it does not seem possible to meet demands age through traditional manual reading information extraction gaining knowledge. this paper, workflow is proposed extract prospecting by text mining based on convolutional neural networks (CNNs). The aim classify automatically. procedure involves three parts: 1) acquisition; 2)...

The specific surface area (SSA) of ABO3-type perovskite is one the important properties associated with photocatalytic ability. In this work, data mining methods were used to explore relationship between SSA (in range 1-60 m2 g-1) and its features, including chemical compositions technical parameters. genetic algorithm-support vector regression method was screen main features for modeling. correlation coefficient ( R) predicted experimental SSAs reached as high 0.986 training set 0.935...

With the era of big data, prediction and evaluation geological mineral resources have gradually entered into a new stage from digital prospecting to intelligent prospecting. The theoretical method data mining can contribute deep resource evaluation. This paper extracts ore-causing ore-caused anomaly information based on text technology, constructs regional conceptual model data. First, we set up corpus discovery preprocessing technology. Second, used CNN multiple scale classification...

Predicting future blood glucose (BG) levels in diabetic patients can help reduce high and low events. The prediction method this article is as follows: the improved grey wolf optimizer combined with kernel extreme learning machine, called I-GWO-KELM. data set of was obtained by continuous monitoring. Initially, Kalman filter used to errors during collection. algorithm optimizes parameters including regularization coefficients parameters. Then KELM predict glucose, mean absolute error (MAE)...

Census transform has been identified to be very effective and robust for stereo matching. But it needs too much computation the accuracy is limited. In this paper, a semi-global matching based on modified census proposed decrease computation. To take use of parallel like CUDA, new cost model by using gradient maps. The method performs better than original in experiment. Compared method, reduces time consumption while remaining accuracy.

Recently, patent data analysis has attracted a lot of attention, and keyword extraction is hot problem. Most existing methods for are based on the frequency words without semantic information. In this paper, we propose an Unsupervised Keyword Extraction Method (UKEM) Chinese clustering. More specifically, use Skip-gram model to train word embeddings corpus. Then each represented as vector called vector. These vectors clustered obtain several cluster centroids. Next, distance between in...

Active learning aims to select the most valuable unlabelled samples for annotation. In this paper, we propose a redundancy removal adversarial active (RRAAL) method based on norm online uncertainty indicator, which selects their distribution, uncertainty, and redundancy. RRAAL includes representation generator, state discriminator, module (RRM). The purpose of generator is learn feature sample, discriminator predicts vector after concatenation. We added sample improve ability designed...

This paper is main to find the best way make a plan about generators operation. Firstly, we describe all conditions as constraints exactly. Secondly, simplify models by using zero-one liner programming. Thirdly, use non-liner optimization and integer programming create objective function. Because of great quantity work, interior-point method an example increase efficiency. Also, offer get optimum solution Lingo. The solving process provides theories analysis foundation for other related problems.

A new graphical model called asymmetric multi-agent influence diagrams (AMAEDs) is proposed in the paper, which based on (MAIDs) and (AIDs). Firstly, we indicate that using MAIDs to represent games may cause representation explosion inefficient solving. Then, AMAIDs deal with this problem, an example introduced show application of AMAIDs.

This paper proposes a China Railway Parcel traffic amount forecast model. The model can effectively solve the variable multiple correlation problem in regression equation. Based on partial least squares regression, prediction of Traffic Volume was calculated. problem.

While many approaches have been proposed to analyze the problem of matrix multiplication parallel computing, few them address on heterogeneous processor platforms. It still remains an open question platforms find optimal schedule that balances load within set while minimizing amount communication. A great studies are based rectangular partition, whereas optimality partition as basis has not well justified. In this paper, we propose a new method schedules with mixed co-design goal total...

In this paper we introduce and study the online consistent $k$-clustering with outliers problem, generalizing non-outlier version of problem studied in [Lattanzi-Vassilvitskii, ICML17]. We show that a simple local-search based algorithm can give bicriteria constant approximation for $O(k^2 \log^2 (nD))$ swaps medians (recourse) total, where $D$ is diameter metric. When restricted to without outliers, our simpler, deterministic gives better ratio recourse, compared