Evandro Marcos Saidel Ribeiro

Application of the method of continued fractions to electron scattering by polyatomic molecules

AbstractA new computational code based on the method of continued fractions is developed to study low-energy electron scatteringby polyatomic molecules with arbitrary symmetry. As a first application, this code is used to study elastic electron scatteringby four representative molecules with different symmetries, viz., H 2 ,CH 4 ,H 2 O, and NH 3 . The calculated physical quantitiessuch as reactance K matrices, differential cross sections, etc, converge rapidly. Also, our calculated cross sections are in goodagreement with other available theoretical and/or experimental data, confirming that this is an efficient method for studyingelectron molecule scattering.  2001 Elsevier Science B.V. All rights reserved. PACS: 34.80.Bm; 02.70.-c; 03.65.NkKeywords: Polyatomic molecules; Electron scattering; H 2 ;CH 4 ;H 2 O; NH 3 ; Cross sections 1. IntroductionThe application of the method of continued frac-tions(MCF) to electron-atomscatteringwas originallyproposed by Horaˇcek and Sasakawa in 1983 [1]. Intheir paper, the

Scattering lengths for Li–Cs, Na–Cs, K–Cs, and Rb–Cs ultracold collisions

We present triplet and singlet scattering lengths for X--133Cs collisions (X=6,7Li, 23Na, 39,40,41K, and 85,87Rb). We consider the short-range potential-energy curves based on high-quality ab initio calculations and the long-range potential described by accurate dispersion coefficients available in the literature. Using a Fermi function to smoothly connect the short- and long-range terms we construct the potential for all R and evaluate the scattering length for each mixed species X--Cs. In particular, we predict that the cross section for inelastic processes between 23Na and 133Cs and between 40K and 133Cs will be small as compared with the respective elastic cross sections.

Extension of the method of continued fractions to molecular photoionization: an application to ammonia

A computational code based on the method of continued fractions, previously developed by our group for electron–molecule scattering calculations, is extended to treat photoionization of molecules of arbitrary symmetry. This new computational code is applied to study the photoionization of the two outermost valence orbitals of ammonia in the exact static-exchange level of approximation. The method has proved to be very accurate and rapidly convergent. Our results obtained for cross sections agree well with both experimental and theoretical results available in the literature.

Information theoretic approach for accounting classification

In this paper we consider an information theoretic approach for the accounting classification process. We propose a matrix formalism and an algorithm for calculations of information theoretic measures associated to accounting classification. The formalism may be useful for further generalizations and computer-based implementation. Information theoretic measures, mutual information and symmetric uncertainty, were evaluated for daily transactions recorded in the chart of accounts of a small company during two years. Variation in the information measures due the aggregation of data in the process of accounting classification is observed. In particular, the symmetric uncertainty seems to be a useful parameter for comparing companies over time or in different sectors or different accounting choices and standards.

A theoretical procedure to treat elastic electron-molecule collision using the configuration-interaction method

A procedure to use a configuration-interaction (CI) wavefunction of the target to determine the potential interaction in the electron-molecule collision theory is proposed. Static, exchange and polarization contributions are obtained explicitly. As a first application the method is used to study elastic scattering by two molecules with different symmetries, H2 and CH4. Differential cross sections (DCS) using Hartree-Fock (HF) and CI wavefunctions are presented and the effect of electron correlation analysed. The DCS were obtained using the Schwinger variational iterative method and the method of continued fractions. It is observed that the discrepancies between HF and CI results are more significant for low energies and small angles. Although our procedure is valid for arbitrary energies, it could be most useful for electron-molecule collision in the low-energy range.

Desempenho financeiro como critério de decisão para seleção modelos de análise e concessão de crédito

This paper aims to show the importance of the use of financial metrics in decision-making of credit scoring models selection. In order to achieve such, we considered an automatic approval system approach and we carried out a performance analysis of the financial metrics on the theoretical portfolios generated by seven credit scoring models based on main statistical learning techniques. The models were estimated on German Credit dataset and the results were analyzed based on four metrics: total accuracy, error cost, risk adjusted return on capital and Sharpe index. The results show that total accuracy, widely used as a criterion for selecting credit scoring models, is unable to select the most profitable model for the company, indicating the need to incorporate financial metrics into the credit scoring model selection process.

Aprendizagem estatística aplicada à previsão de default de crédito

OBJETIVO Esta pesquisa tem por objetivo comparar a capacidade preditiva das principais tecnicas de aprendizagem estatistica atualmente utilizadas nacional e internacionalmente para tratamento do problema de risco de credito, analisando-as a luz de sua eficiencia preditiva. METODOLOGIA No presente estudo foram utilizados os dados da base German Credit Dataset. As estimacoes foram obtidas a partir do pacote e1071 aplicado ao software estatistico R®. Foram testadas, comparativamente, as tecnicas: Analise Discriminante, Regressao Logistica, Redes Bayesianas Naive Bayes, kdB-1, kdB-2, SVC e SVM. Como ponto de corte para Analise Discriminante, Regressao Logistica, Redes Bayesianas Naive Bayes, kdB-1, kdB-2 foi empregada a curva ROC - Receiver Operating Characteristic. Os resultados foram comparados com base na acuracia (medida pelo metodo ACC) e pelo criterio de custo pelo erro (medido com base nos criterios trazidos pela base de dados) e os resultados foram apresentados sob o formato de uma matriz de confusao e validados pelo metodo k-Fold Cross-Validation. RESULTADOS E CONCLUSOES Analisando os resultados das tecnicas o SVM apresentou o maior numero medio de acertos, obtendo indice de acuracia 86.6%, bem como as maiores medias de acerto em termos de previsao dentre todos os modelos, com aproximadamente 85,5% de acertos na previsao de default. Pela analise do custo total e possivel verificar que o custo do erro do SVM foi de 50.46, sendo que 41.44% desse custo foi produzido por erros de falso positivo e 58.56% do custo foi produzido por erro de falso negativo. Este foi o menor custo dentre os custos produzidos pelas tecnicas estudadas, sendo aproximadamente 42% menor que o segundo colocado na analise. Para ambas as metricas (ACC e custos do erro) observou-se a superioridade do metodo SVM – Support Vector Machine, sugerindo que este metodo apresenta caracteristicas que potencializam o seu poder preditivo para o caso estudado. Tais resultados corroboram com indicacoes de autores que mostram que a superioridade do SVM se deve ao fato de que metodos ensemble, como e o caso do SVM, em dados desbalanceados. IMPLICACOES PRATICAS Com base nos resultados e possivel verificar que a evolucao das tecnicas de aprendizagem estatistica tem contribuido para a maximizacao dos resultados de classificacao, sendo este o problema central da analise de credito. Pela quantidade de estudos realizados internacionalmente e destacados na bibliografia, este se mostra um campo fertil para o desenvolvimento de pesquisas futuras, especialmente para comparacao dos resultados alcancados por este estudo. Especificamente sobre o SVM, os resultados corroboram com estudos que indicam boa aderencia de tal modelo para situacoes reais nas quais o banco de dados a ser classificado e desbalanceado. PALAVRAS-CHAVE Risco de credito, Aprendizagem estatistica, previsao de default.   STATISTICAL LEARNING APPLIED TO CREDIT DEFAULT PREDICTION OBJECTIVE This research compares the predictive potential of the main statistical learning techniques currently used in both nationally and internationally extent for the treatment of credit risk problem, by analyzing their predictive efficiency. METHODOLOGY In this study we have used the data base nominated German Credit Dataset. The estimates were obtained from e1071 package applied in R®. The comparatively seven techniques were tested: linear discriminant analysis, logistic regression, Naive Bayes, KDB-1, KDB-2, SVC and SVM. To discriminant analysis, logistic regression, Naive Bayes, KDB-1, KDB-2 employed the ROC curve - Receiver Operating Characteristic as a cut-off point. The results were compared based on accuracy (measured by the ACC technique) and on error cost criterion (measured by criteria observed in the database studied) and the results were presented in the confusion matrix format and validated by the k-fold Cross Validation technique. RESULTS AND CONCLUSIONS The results showed the highest average number of SVM technique hits, getting the highest accuracy (86.6%), as well as the biggest hits in terms of average forecast among all the other models with approximately 85.5% of correct default prediction. By the total cost results, we can verify that the SVM error cost was 50.46, with 41.44 percent of this cost produced by false positive errors and 58.56% by false negative error. It was the lowest cost among the costs produced by the studied techniques, approximately 42% lower than the second one. For both metrics (ACC and error costs) this study observed superiority of SVM method suggesting this method has characteristics that enhance the predictive power for the case studied. These results corroborate with results of authors that shows the SVM’s superiority is due to the ensemble methods such as SVM, by their nature, are more effective for unbalanced data. PRACTICAL IMPLICATIONS Based on the results is possible to check that the evolution of the statistical learning techniques have contributed to the maximization of classification results, being this the central problem of credit analysis. The amount of studies conducted internationally and showed in the bibliographical study, this is a fertile field to the future research development, especially to compare the results obtained by this study. About the SVM method, the results corroborate studies that indicate good adhesion of such model to real situations in which the dataset’s unbalanced. KEYWORDS Credit Risk, Statistical Learning, Default Prediction.

A theoretical study on the photoionization of the valence orbitals of phosphine

We report a theoretical study on the photoionization of phosphine in the static-exchange level and frozen core approximation, using the method of continued fractions. The main subject of the present study is to investigate in which extent the Hartree-Fock description of the target applied to molecular photoionization is valid. Also, the role played by multichannel coupling is analysed. Our study shows that single-channel Hartree-Fock calculations can provide reliable results except for photon energies near the photoionization threshold.