Support Vector Machines (SVM) is a practical algorithm that has been widely used in many areas. To guarantee its satisfying performance, it is important to set appropriate parameters of SVM algorithm. Sequential Minimal Optimization (SMO) is an effective training algorithm belonging to SVM, so is LS_SVM....

This paper presents the scheduling problem with multi-resource sharing, which each operation may need more than one kinds of resource. Timed Petri net is used to formulate this problem to analyze deadlock and minimize the makespan. A deadlock avoidance policy addressed here consists of three stages:...

As to the problem that normal SVM algorithm has a high computational complexity with large scale data and the method of selecting parameters of the study machine is complexity,we improved the SMO algorithm in two aspects of structure and parametric selection to increase operational speed and efficiency...

A new algorithm is proposed, which is called Generalized Ant Colony Optimization (GACO) algorithm. Two new functions are presented to model the behavior for describing the pheromone evaporation and pheromone added to the edges that belong to the best-so-far solution. A class of strictly increasing function...

Visualizing hierarchy of a website is very helpful for both users’ navigating and search engine efficiently presenting results. In this paper, treating webpages as nodes and hyperlinks as directed edges, the link structure is firstly modeled as weighted directed graph. Considering multiple website features,...

This paper presents a phrase combination approach to patent SMT (Statistical Ma-chine Translation) for Japanese to English. To minimize the segmentation problems caused by the rich OOV (out-of-vocabulary) words in the patent texts, the character based translation phrases are first introduced to avoid...

This paper presents a new Chinese chunking algorithm based on cascaded conditional random fields. Conditional random fields solve the tagging problems well, while the cascaded models restrain the affection of part-of-speech errors. The experimental results show that this approach achieves impressive...

This paper proposes a new part-of-speech tagging algorithm based on the fusion model which combines Maximum Entro-py model and Error Correction model. According to the analysis of the two models, the fusion tagging model is uti-lized with the profits of conditional prob-ability model and rule based model....

Growing convergence among mobile computing devices and embedded tech-nology promotes the development and the deployment of “context-aware” applications, where location is the most essential context. In this paper we present an improved probabilistic method based on RFID (Radio Frequency Identification)...

Web information retrieval technology has been widespread concerned by research-ers. Web-based search of the OOV Trans-lation Mining has also become hot spots. In this paper, the re-ordering of OOV translation candidates is studied, which is the result of web mining. Automatic word alignment technology...

In [1], Fung et al, had constructed by a very fast algorithm: PSVM classifier, which mainly makes use of the Sherman-Morrison-Woodbury (SWM) identity [1, 7, 8]. However, for one thing, when handling nonlinear problems, the matrix in (1) always is of dimension , such that the SWM identity is of no use....

Self-selective routing (SSR) protocol is a novel self-selective, self-maintained and fault-tolerant routing protocol for MANET. SSR has some fatal faults, such as high packet collision and only providing low available andwidth. In this paper, we present a new routing protocol called Self-selection Based...

There has been considerable interest in development of scholar search engines. A disadvantage of current scholar search engines is that they have not explored the relationship among theses fully. A novel scholar search engine is proposed here to fill this gap, which can detect different topics hidden...

Mixed multi-unit combinatorial auctions provide two types of bids as XOR-bids and OR-bids, which make them popular for resource allocation in marketplace. However, the problem of maximizing the revenue of auctioneers is a NP-complete problem. In this paper, we propose an improved partheno-genetic algorithm...