Machine learning and artificial intelligence have achieved a human-level performance in many application domains, including image classification, speech recognition and machine translation. However, in the financial domain expert-based credit risk models have still been dominating. Establishing meaningful benchmark and comparisons on machine-learning approaches and human expert-based models is a prerequisite in further introducing novel methods. Therefore, our main goal in this study is to establish a new benchmark using real consumer data and to provide machine-learning approaches that can serve as a baseline on this benchmark. We performed an extensive comparison between the machine-learning approaches and a human expert-based model—FICO credit scoring system—by using a Survey of Consumer Finances (SCF) data. As the SCF data is non-synthetic and consists of a large number of real variables, we applied two variable-selection methods: the first method used hypothesis tests, correlation and random forest-based feature importance measures and the second method was only a random forest-based new approach (NAP), to select the best representative features for effective modelling and to compare them. We then built regression models based on various machine-learning algorithms ranging from logistic regression and support vector machines to an ensemble of gradient boosted trees and deep neural networks. Our results demonstrated that if lending institutions in the 2001s had used their own credit scoring model constructed by machine-learning methods explored in this study, their expected credit losses would have been lower, and they would be more sustainable. In addition, the deep neural networks and XGBoost algorithms trained on the subset selected by NAP achieve the highest area under the curve (AUC) and accuracy, respectively.

Pairwise relationships extracted from biomedical literature are insufficient in formulating biomolecular interactions. Extraction of complex relations (namely, biomedical events) has become the main focus of the text-mining community. However, there are two critical issues that are seldom dealt with by existing systems. First, an annotated corpus for training a prediction model is highly imbalanced. Second, supervised models trained on only a single annotated corpus can limit system performance. Fortunately, there is a large pool of unlabeled data containing much of the domain background that one can exploit.

This paper determined the predefining factors of loan repayment behavior based on psychological and behavioral economics theories. The purpose of this research is to identify whether an individual’s credit risk can be predicted based on psychometric tests measuring areas of psychological factors such as effective economic decision-making, self-control, conscientiousness, selflessness and a giving attitude, neuroticism, and attitude toward money. In addition, we compared the psychological indicators to the financial indicators, and different age and gender groups, to assess whether the former can predict loan default prospects. This research covered the psychometric test results, financial information, and loan default information of 1118 borrowers from loan-issuing applications on mobile phones. We validated the questionnaire using confirmatory factor analysis (CFA) and achieved an overall Cronbach’s alpha reliability coefficient greater than 0.90 (α = 0.937). We applied the empirical data to construct prediction models using logistic regression. Logistic regression was employed to estimate the parameters of a logistic model. The outcome indicates that positive results from the psychometric testing of effective financial decision-making, self-control, conscientiousness, selflessness and a giving attitude, and attitude toward money enable individuals’ debt access possibilities. On the other hand, one of the variables—neuroticism—was determined to be insignificant. Finally, the model only used psychological variables proven to have significant default predictability, and psychological variables and psychometric credit scoring offer the best prediction capacities.

Credit scoring is a process of determining whether a borrower is successful or unsuccessful in repaying a loan using borrowers’ qualitative and quantitative characteristics. In recent years, machine learning algorithms have become widely studied in the development of credit scoring models. Although efficiently classifying good and bad borrowers is a core objective of the credit scoring model, there is still a need for the model that can explain the relationship between input and output. In this work, we propose a novel partially interpretable adaptive softmax (PIA-Soft) regression model to achieve both state-of-the-art predictive performance and marginally interpretation between input and output. We augment softmax regression by neural networks to make it adaptive for each borrower. Our PIA-Soft model consists of two main components: linear (softmax regression) and non-linear (neural network). The linear part explains the fundamental relationship between input and output variables. The non-linear part serves to improve the prediction performance by identifying the non-linear relationship between features for each borrower. The experimental result on public benchmark datasets shows that our proposed model not only outperformed the machine learning baselines but also showed the explanations that logically related to the real-world.

One essential task in extracting information from biomedical literature is the bio Named Entity Recognition (NER) process, which basically defines the boundaries between typical words and biomedical terminology in particular text data, and assigns them based on domain knowledge. This paper presents a semi supervised integration of completely different classifiers to cover knowledge from unlabeled data to recognize bio named entities in text. We modified the original co-training, a semi supervised learning algorithm, with a scalable feature processing schema, which extracts the bio NER feature from a number of unlabeled data and converts different types of feature sets. Our base result shows that the classifiers of co-training achieve significant learning from unlabeled data.