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.     

A study of standardization of variables in cluster analysis

A methodological problem in applied clustering involves the decision of whether or not to standardize the input variables prior to the computation of a Euclidean distance dissimilarity measure. Existing results have been mixed with some studies recommending standardization and others suggesting that it may not be desirable. The existence of numerous approaches to standardization complicates the decision process. The present simulation study examined the standardization problem. A variety of data structures were generated which varied the intercluster spacing and the scales for the variables. The data sets were examined in four different types of error environments. These involved error free data, error perturbed distances, inclusion of outliers, and the addition of random noise dimensions. Recovery of true cluster structure as found by four clustering methods was measured at the correct partition level and at reduced levels of coverage. Results for eight standardization strategies are presented. It was found that those approaches which standardize by division by the range of the variable gave consistently superior recovery of the underlying cluster structure. The result held over different error conditions, separation distances, clustering methods, and coverage levels. The traditionalz-score transformation was found to be less effective in several situations.     

Alzheimer’s disease: the impact of age-related changes in reproductive hormones

No Abstract.  .     

Phosphodiesterase: overview of protein structures, potential therapeutic applications and recent progress in drug development

Phosphodiesterases (PDEs) are essential regulators of cyclic nucleotide signaling with diverse physiological functions. Because of their great market potential and therapeutic importance, PDE inhibitors became recognized as important therapeutic agents in the treatment of various diseases. Currently, there are seven PDE inhibitors on the market, and the pharmacological and safety evaluations of many drug candidates are in progress. Three-dimensional (3D) structures of catalytic domains of PDE 1, -3, -4, -5 and -9 in the presence of their inhibitors are now available, and can be utilized for rational drug design. Recent advances in molecular pharmacology of PDE isoenzymes resulted in identification of new potential applications of PDE inhibitors in various therapeutic areas, including dementia, depression and schizophrenia. This review will describe the latest advances in PDE research on 3D structural studies, the potential of therapeutic applications and the development of drug candidates.Received 30 November 2004; received after revision 24 January 2005; accepted 5 February 2005     

Decision-making at the surface of the intact or barrier disrupted skin: potential applications for vaccination or therapy

The skin is a highly accessible organ and constitutes an active immunological site. Both these properties make this surface an attractive route for what promises to be a cost-effective, simple, practical and needle-free delivery of vaccines and immunomodulators. Less obvious is the fact that the state of the skin barrier can influence quantitative and qualitative aspects of antigen-specific immune responses. The everyday decision-making at the skin epithelium concerns the choice between the induction of an immune response and the establishment of a state of non-responsiveness (tolerance). This decision is influenced by various factors such as the dose, the route (intact vs barrier-disrupted skin), the cytokine microenvironment and the nature of the antigenic stimulus. By increasing our understanding of how immune responses are regulated in the epidermis we can envisage the development of immunisation protocols aimed at eliciting a protective immune response or inducing tolerance, with direct applications to preventive or therapeutic vaccination, respectively.Received 29 November 2004; received after revision 2 February 2005; accepted 22 February 2005     

Marrow mesenchymal stem cells transduced with TPO/FL genes as support for ex vivo expansion of hematopoietic stem/progenitor cells

A new marrow-derived mesenchymal stem cell (hMSC) line that could support expansion of hematopoietic stem/progenitor cells (HSPCs) was developed. Primary hMSCs were infected with retrovirus containing Flt-3 ligand and thrombopoietin genes. CD34+ cells from cord blood were expanded with primary hMSCs or transduced hMSCs. The expansion of total nucleated cells, CD34+ cells and mixed colonies containing erythroid and myeloid cells and megakaryocytes for 2 weeks coculture with transduced hMSCs was remarkably increased. The outputs of long-term culture-initiating cells for 2 and 4 weeks coculture with transduced hMSCs were also largely increased. The expansion rates of HSPCs with transduced hMSCs were unchanged for 6 weeks. In contrast, the expansion rates of HSPCs with primary hMSCs declined drastically through 6 weeks. SCID-repopulating cell expansion with transduced hMSCs for 4 weeks was significantly higher than that of uncultured CD34+ cells and HSPCs expanded with primary hMSCs. Received 21 June 2005; received after revision 30 July 2005; accepted 24 August 2005     

Congenital muscular dystrophy: molecular and cellular aspects

The congenital muscular dystrophies are a clinically and genetically heterogeneous group of neuromuscular disorders. Each form has a characteristic phenotype, but there is overlap between some entities and their classification is based on a combination of clinical features and the primary or secondary protein defect. Recent studies have identified the genetic basis of a number of congenital muscular dystrophies (11 genes in total) and have recognised a novel pathological mechanism that highlights the importance of the correct posttranslational processing of proteins, in particular -dystroglycan. Diagnosis of these conditions has been aided by the availability of specific antibodies for each protein and a better understanding of the protein changes that accompany each condition. In this review we present the major molecular, clinical and diagnostic aspects of each group of congenital muscular dystrophy with an emphasis in the more recent developments.Received 11 December 2004; accepted 15 December 2004     

Modeling the MHC class I pathway by combining predictions of proteasomal cleavage,TAP transport and MHC class I binding

Epitopes presented by major histocompatibility complex (MHC) class I molecules are selected by a multi-step process. Here we present the first computational prediction of this process based on in vitro experiments characterizing proteasomal cleavage, transport by the transporter associated with antigen processing (TAP) and MHC class I binding. Our novel prediction method for proteasomal cleavages outperforms existing methods when tested on in vitro cleavage data. The analysis of our predictions for a new dataset consisting of 390 endogenously processed MHC class I ligands from cells with known proteasome composition shows that the immunological advantage of switching from constitutive to immunoproteasomes is mainly to suppress the creation of peptides in the cytosol that TAP cannot transport. Furthermore, we show that proteasomes are unlikely to generate MHC class I ligands with a C-terminal lysine residue, suggesting processing of these ligands by a different protease that may be tripeptidyl-peptidase II (TPPII).Received 26 November 2004; received after revision 4 February 2005; accepted 4 March 2005S. Tenzer and B. Peters contributed equally to this work.     

Emodin inhibits tumor cell migration through suppression of the phosphatidylinositol 3-kinase-Cdc42/Rac1 pathway

Enhanced cell migration is one of the underlying mechanisms in cancer invasion and metastasis. Therefore, inhibition of cell migration is considered to be an effective strategy for prevention of cancer metastasis. We found that emodin (3-methyl-1,6,8-trihydroxyanthraquinone), an active component from the rhizome of Rheum palmatum, significantly inhibited epidermal growth factor (EGF)- induced migration in various human cancer cell lines. In the search for the underlying molecular mechanisms, we demonstrated that phosphatidylinositol 3-kinase (PI3K) serves as the molecular target for emodin. In addition, emodin markedly suppressed EGF-induced activation of Cdc42 and Rac1 and the corresponding cytoskeleton changes. Moreover, emodin, but not LY294002, was able to block cell migration in cells transfected with constitutively active (CA)-Cdc42 and CA-Rac1 by interference with the formation of Cdc42/Rac1 and the p21-activated kinase complex. Taken together, data from this study suggest that emodin inhibits human cancer cell migration by suppressing the PI3K-Cdc42/Rac1 signaling pathway.Received 7 February 2005; received after revision 11 March 2005; accepted 18 March 2005