A multiple intelligent agent system for credit risk prediction via an optimization of localized generalization error with diversity

Company bankruptcies cost billions of dollars in losses to banks each year. Thus credit risk prediction is a critical part of a bank＇s loan approval decision process. Traditional financial models for credit risk prediction are no longer adequate for describing today＇s complex relationship between the financial health and potential bankruptcy of a company. In this work, a multiple classifier system （embedded in a multiple intelligent agent system） is proposed to predict the financial health of a company. In our model, each individual agent （classifier） makes a prediction on the likelihood of credit risk based on only partial information of the company. Each of the agents is an expert, but has limited knowledge （represented by features） about the company. The decisions of all agents are combined together to form a final credit risk prediction. Experiments show that our model out-performs other existing methods using the benchmarking Compustat American Corporations dataset.     

ROLE OF α-PSEUDO-UNIVEX FUNCTIONS IN VECTOR VARIATIONAL-LIKE INEQUALITY PROBLEMS

In this paper, we introduce a new class of generalized convex function, namely, a-pseudounivex function, by combining the concepts of pseudo-univex and α-invex functions. Further, we establish some relationships between vector variational-like inequality problems and vector optimization problems under the assumptions of α-pseudo-univex functions. Results obtained in this paper present a refinement and improvement of previously known results.     

Character and OTU stability in five taxonomic groups

The character and OTU stability of classifications based on UPGMA clustering and maximum parsimony (MP) trees were compared for 5 datasets (families of angiosperms, families of orthopteroid insects, species of the fish genusIctalurus, genera of the salamander family Salamandridae, and genera of the frog family Myobatrachidae). Stability was investigated by taking different sized random subsamples of OTUs or characters, computing UPGMA clusters and an MP tree, and then comparing the resulting trees with those based on the entire dataset. Agreement was measured by two consensus indices, that of Colless, computed from strict consensus trees, and Stinebrickner's 0.5-consensus index. Tests of character stability generally showed a monotone decrease in agreement with the standard as smaller sets of characters are considered. The relative success of the two methods depended upon the dataset. Tests of OTU stability showed a monotone decrease in agreement for UPGMA as smaller sets of OTUs are considered. But for MP, agreement decreased and then increased again on the same scale. The apparent superiority of UPGMA relative to MP with respect to OTU stability depended upon the dataset. Considerations other than stability, such as computer efficiency or accuracy, will also determine the method of choice for classifications.     

广西产眼镜王蛇毒α-神经毒素基因的分子克隆及序列分析

为获得眼镜王蛇（Ophiophagus hannah，简称Oh）蛇毒α－神经毒素（α-NT）的基因序列，依据眼镜蛇科不同毒蛇种类来源的α－NT基因有较高的同源性，设计1对上下游引物，为克服引物带来模糊扩增，在蛋白编码部分再设计1对上下游特异引物，用Nacleospin RNA Kit法从3条活眼镜王蛇毒腺中提取mRNA，以3′端引物合成的cDNA作为模板进行PCR扩增反应，测定产物的核苷酸序列，得到全长474bp的眼镜王蛇cDNA基因核苷酸序列。该核苷酸序列的信号肽与眼镜蛇树属Pseudonnaja textilis(Pt)、海蛇Laticauda semifasciata(Ls)100%同源，与眼镜蛇南洋亚种Naja sputatrix (Ns)、银环蛇（Bungarus multicinctus)(Bm)96.8%同源；蛋白密码部分有83．3％与Ns、79．2％与Pt、76．4％与Ls、74．1％与Bm同源。信号肽后紧接着的72个氨基酸有90．3％与已发现的眼镜王蛇毒长链α-NT Toxin a同源，大约有73．6％与Toxin b、69．7％与Oh-4、66．7％与Oh-5、56．9％与Oh-6A和6B同源，并与α－银环蛇毒素54．2％同源。说明新发现的眼镜王蛇cDNA是一条长链α-NT基因。     

A relevant in vitro rat model for the evaluation of blood-brain barrier translocation of nanoparticles

Poly(MePEG₂₀₀₀cyanoacrylate-co-hexadecylcyanoacrylate) (PEG-PHDCA) nanoparticles have demonstrated their capacity to reach the rat central nervous system after intravenous injection. For insight into the transport of colloidal systems across the blood-brain barrier (BBB), we developed a relevant in vitro rat BBB model consisting of a coculture of rat brain endothelial cells (RBECs) and rat astrocytes. The RBECs used in our model displayed and retained structural characteristics of brain endothelial cells, such as expression of P-glycoprotein, occludin and ZO-1, and immunofluorescence studies showed the specific localization of occludin and ZO1. The high values of transendothelial electrical resistance and low permeability coefficients of marker molecules demonstrated the functionality of this model. The comparative passage of polyhexadecylcyanoacrylate and PEG-PHDCA nanoparticles through this model was investigated, showing a higher passage of PEGylated nanoparticles, presumably by endocytosis. This result was confirmed by confocal microscopy. Thanks to a good in vitro/in vivo correlation, this rat BBB model will help in understanding the mechanisms of nanoparticle translocation and in designing new types of colloidal carriers as brain delivery systems.Received 4 March 2005; accepted 14 April 2005     

Protein kinase C and phospholipase D: intimate interactions in intracellular signaling

Diacylglycerol (DAG) was discovered as a potent lipid second messenger with protein kinase C (PKC) as its major cellular target more than 25 years ago. There is increasing evidence of significant complexity within lipid signaling, and the classical DAG-PKC model no longer stands alone but is part of a larger bioactive lipid universe involving glycerolipids and sphingolipids. Multiple layers of regulation exist among PKC- and DAG-metabolizing enzymes such as phosphatidylcholine (PC)-specific phospholipase D, and cross-talk exists between the glycerolipid and sphingolipid pathways, with PKC at the center. Currently, there is intense interest in the question of whether DAG derived from PC can function as a lipid second messenger and regulate PKC analogous to DAG derived from phosphatidylinositol-4,5-bisphosphate (PIP₂). To address these issues and incorporate DAG-PKC and other signaling pathways into an expanded view of cell biology, it will be necessary to go beyond the classical approaches and concepts.Received 29 November 2004; received after revision 18 January 2005; accepted 4 March 2005This work is dedicated to the memory of Dr. Yasutomi Nishizuka, the discoverer of protein kinase C, who was both a gentleman and a scientist.     

Nucleotide analogues as probes for DNA polymerases

Transmission of the genetic information from the parental DNA strand to the offspring is crucial for the survival of any living species. In nature, all DNA synthesis in DNA replication, recombination and repair is catalyzed by DNA polymerases and depends on their ability to select the canonical nucleobase pair from a pool of structurally similar building blocks. Recently, a wealth of valuable new insights into DNA polymerase mechanisms have been gained through application of carefully designed synthetic nucleotides and oligonucleotides in functional enzyme studies. The applied analogues exhibit features that differ in certain aspects from their natural counterparts and, thus, allow investigation of the involvement and efficacy of a chosen particular aspect on the entire complex enzyme mechanism. This review will focus on a depiction of the efforts that have been undertaken towards the development of nucleotide analogues with carefully altered properties. The different approaches will be discussed in the context of the motivation and the problem under investigation.Received 16 March 2005; received after revision 5 May 2005; accepted 8 June 2005     

Algorithmic compression of empirical data: reply to Twardy, Gardner, and Dowe

This discussion note responds to objections by Twardy, Gardner, and Dowe to my earlier claim that empirical data sets are algorithmically incompressible. Twardy, Gardner, and Dowe hold that many empirical data sets are compressible by Minimum Message Length technique and offer this as evidence that these data sets are algorithmically compressible. I reply that the compression achieved by Minimum Message Length technique is different from algorithmic compression. I conclude that Twardy, Gardner, and Dowe fail to establish that empirical data sets are algorithmically compressible.