No abstract

Credit scoring has become a critical and challenging management science issue as the credit industry has been facing stiffer competition in recent years. Many classification methods have been suggested to tackle this problem in the literature. In this paper, we investigate the performance of various credit scoring models and the corresponding credit risk cost for three real-life credit scoring data sets. Besides the well-known classification algorithms (e.g. linear discriminant analysis, logistic regression, neural networks and k-nearest neighbor), we also investigate the suitability and performance of some recently proposed, advanced data mining techniques such as support vector machines (SVMs), classification and regression tree (CART), and multivariate adaptive regression splines (MARS). The performance is assessed by using the classification accuracy and cost of credit scoring errors. The experiment results show that SVM, MARS, logistic regression and neural networks yield a very good performance. However, CART and MARS's explanatory capability outperforms the other methods.

We introduced Eu2+ ions as the electron donor to provide a reducing environment to suppress the oxidation of Sn2+. Finally, we obtained a pure red (PEA)2SnI4 LED with a high EQE up to 1.48% and a maximum brightness of 221 cd m−2.

Elastic strains in metallic catalysts induce enhanced selectivity for carbon dioxide reduction (CO2R) toward valuable multicarbon (C2+) products. However, under working conditions, the structure of catalysts inevitably undergoes reconstruction, hardly retaining the initial strain. Herein, we present a metal/metal oxide synthetic strategy to introduce and maintain the tensile strain in a copper/ceria heterostructure, enabled by the presence of a thin interface layer of Cu2O/CeO2. The tensile strain in the copper domain and deficient electron environment around interfacial Cu sites resulted in strengthened adsorption of carbonaceous intermediates and promoted *CO dimerization. The strain effect in the copper/ceria heterostructure leads to an improved C2+ selectivity with a maximum Faradaic efficiency of 76.4% and a half-cell power conversion efficiency of 49.1%. The fundamental insights gained from this system can facilitate the rational design of heterostructure catalysts for CO2R.

Lithium-ion capacitors (LICs), consisting of a battery-like negative electrode and a capacitive porous-carbon positive electrode, deliver more than twice the energy density of electric double-layer capacitors. However, their wide application suffers from low energy density and reduced cycle life at high rates. Herein, hierarchical meso-microporous carbon nanospheres with a highly disordered structure and nitrogen/phosphorous co-doped properties were synthesized through a facile template method. Such hierarchical porous structure facilitates rapid ion transport, and the highly disordered structure and high heteroatom content provide abundant active sites for Li + charge storage. Electrochemical experiments demonstrated that the carbon nanosphere anode delivers large reversible capability, greatly improves rate capability and exhibits excellent cycle stability. An LIC fabricated with the carbon nanosphere anode and an activated carbon cathode yields a high energy density of 103 W h kg −1 , an extremely high power density of 44,630 W kg −1 , and longterm cyclability of over 10,000 cycles. This work presents how structural control of carbon materials at the nano/atomic scale can significantly enhance electrochemical performance, enabling new opportunities for the design of high-performance energy-storage devices.