Predicting the occurrence of a particular event interest at future time points is primary goal survival analysis. The presence incomplete observations due to limitations or loss data traces known as censoring which brings unique challenges in this domain and differentiates analysis from other standard regression methods. popularly used methods such Cox proportional hazard model parametric suffer some strict assumptions hypotheses that are not realistic most real-world applications. To...

Dialogue system is an important application of natural language processing in human-computer interaction. Emotion analysis dialogue aims to classify the emotion each utterance dialogue, which crucially system. In system, helpful semantic understanding and response generation great significance practical customer service quality inspection, intelligent chatbots, so on. However, it challenging solve problems short text, synonyms, neologisms, reversed word order for dialogue. this paper, we...

Abstract Accurate in-hospital length of stay prediction is a vital quality metric for hospital leaders and health policy decision-makers. It assists with decision-making informs operations involving factors such as patient flow, elective cases, human resources allocation, while also informing care risk considerations. The aim the research reported in this paper to use survival analysis model General Internal Medicine (GIM) stay, Shapley value support interpretation resulting model. Survival...

Code reuse attacks have become one of the most popular exploitation techniques, and coarse-grained control flow integrity (CFI) is a practical approach used to prevent such attacks. Recently, some new approaches been proposed construct call-preceded-ROP bypass CFI, however, we find that they still fail shadow stack, which enforces caller-callee semantics strengthen CFI constrains in much stricter way. Therefore, this paper, propose code attack, named loop-oriented programming (LOP), aiming...

In order to build human cognition features into the procedure of clustering, this paper introduces a novel text clustering system, CogTCA (Cognitive Text Clustering with Ants), which (1) represents texts according four cognitive situation dimensions in form matrices and vectors rather than canonical sparse high dimensions, (2) proposes several new similarity measures among texts, (3) implements task as solving combinatorial optimization problem an encounter ant colony algorithm.

The initial alignment model of Strapdown Inertial Navigation System (SINS) with large misalignment angle is a linear and nonlinear mixed system model. Aiming at the model, an improved Cubature Kalman Filter(CKF) algorithm adopted. Firstly, SINS established analyzed. Then, singular value decomposition (SVD) used to assist Cholesky ensure numerical stability in time update phase algorithm, while adjusting modified covariance, performing measurement phase. Finally, experimental results show...

As big data for text has been one of the core types in era artificial intelligence, feature words selection technique become increasingly important analysis. The traditional statistical TF-IDF algorithm lacks semantic information extraction ability text, while network model Textrank applies sentence features to calculation between words. Network such as is very suitable selection, but it does not take influencing factors relationship documents into consideration, so common appearing...

In order to build human cognition features into the procedure of clustering, this paper introduces a novel text clustering system, CogTCA (Cognitive Text Clustering with Ants), which (1) represents texts according four cognitive situation dimensions in form matrices and vectors rather than canonical sparse high dimensions, (2) proposes several new similarity measures among texts, (3) implements task as solving combinatorial optimization problem an encounter ant colony algorithm.

GF-3 series satellites, the first C band and multi-polarization SAR satellite constellation in China, achieve breakthroughs a number of core key technologies. The distribution information architecture with two kinds data buses is used for control, transmission management. In this article, design system net architecture, function housekeeping storage playback flow, important flow payload are described. According to requirements users' convenience, layered automatic mission management applied....

In modern information retrieval systems, the thesaurus is playing an increasingly important role. order to better describe and analyze domain news, this paper proposes a method of keyword extraction, further constructs effective thesaurus. Compared with previous research, grasps core in field by extracting combining keywords, improves effectiveness addition, conducts both manual analysis automated processing construct high-quality thesaurus, which has practical application value. The final...

Topic modeling is a probabilistic generation model to find the representative topic of document and has been successfully applied various document-related tasks in recent years. Especially supervised time model, many methods have achieved some success. The can learn topics from documents annotated with multiple labels that evolve over sequentially organized corpus. However, there are time-stamped reality, which need construct achieve tasks. There few research papers on model. To solve this...

Choquet integral, one of ensemble operators, is used widely in multiple classifiers combination when we consider the interaction between classifiers. The diversity system can affect classification accuracy and generalization ability system. In this paper, two different types classifiers, neural network fuzzy decision tree, are introduced (called mix-combination system). Genetic algorithm to determine a non-additive set function μ which key issue before integral combining Some simulated...

Classification on noisy data streams has recently become one of the most important topics in streaming mining. In this paper, a algorithm for mining Data Streams based Mixture Models C4.5 and NB is proposed called CDSMM. algorithm, used as base classifiers, hypothesis testing method introduced detection concept drifts, Naïve Bayes classifier adopted to filter noise. Extensive experiments demonstrate that CDSMM substantial advantages over similar existing algorithms predictive accuracy with drifts.

This study aims to analyse the behaviour of consumers when they buy books online, and law their scanning products focus check information. In experiment, 20 subjects must follow instruction complete two specific tasks, eye tracker recorded eyes' movement data automatically. According heat map, researchers divided area interest (AOI) which got attention analysed in AOI. Combined with interview analysis, found that: a) The customers tend view pages following an 'F' pattern: is largely focused...

First of all, this paper briefly reviews Nida’s “functional equivalence” theory, and clarifies the core element functional translation that is, response degree target readers to translated text is roughly same as source language readers; Secondly, it introduces characteristics popular science texts; Finally, vocabulary, sentence chapter Takeda Pharmaceutical are selected explain practice in detail, so provide some reference for development translation.

In this study, the research results at home and abroad on basis of semantics, through establishment Russian verb semantic pattern combing significance patterns, to analyze performance syntactic distribution, description Semantic -syntactic as for constructing Machine Translation "verbal semanticsyntactic library", "fixed sentence library" "Variable complete carding three databases achieve a knowledge integration.This study not only can expand category semantics theory, but also promote...

Ancient architectures are one of the most complex structures for surveying and mapping. For lack effective automated mapping method, their have been done manually over years. LIDAR is currently efficient equipment, but limited by aerial platform, its point cloud accuracy fails to meet precision requirement ancient architectures. This paper puts forward a new design flying vehicle, improved in vibration intensity, positioning operating methods. Based on these improvements, can work with...

The three-way decisions strategy has been employed to construct network topology in a single hidden layer feedforward neural (SFNN). However, this model general performance, and does not consider the process costs, since it fixed threshold parameters. Inspired by sequential (STWD), paper proposes STWD with an SFNN (STWD-SFNN) enhance performance of networks on structured datasets. STWD-SFNN adopts multi-granularity levels dynamically learn number nodes from coarse fine, set Specifically, at...