TWD-SFNN: Three-way decisions with a single hidden layer feedforward neural network

Using a genetic backpropagation neural network model for credit risk assessment in the micro, small and medium-sized enterprises

In China, with the "Double Carbon" goal within reach, Micro, Small and Medium-sized Enterprises (MSMEs) emerge as pivotal contributors to economic advancement. However, they are now confronted with the imperative of transitioning towards green and low-carbon practices. To facilitate the attainment of peak carbon dioxide emissions and carbon neutrality, a refined approach is imperative. This entails precise capital allocation, enhanced financial services, streamlined management, and robust risk mitigation strategies. Consequently, conducting thorough credit risk assessments for MSMEs becomes a crucial endeavor. However, obtaining substantial loans for them proves challenging due to their elusive credit ratings and potential defaults. To address this issue, this study leverages machine learning and intelligent optimization algorithms to construct a classification model for default and credit ratings of MSMEs, utilizing their daily invoice data. Specifically, twelve indicators pertaining to default and credit ratings are extracted. Subsequently, Principal Component Analysis is employed to reduce dimensionality and synthesize all pertinent information. Following this, the Genetic Algorithm-based Back Propagation Neural Network (GA-BPNN) is utilized to delineate the relationship between indicators and default, as well as credit rating, respectively. The results indicate a prediction accuracy of 0.92 for default risk and 0.86 for credit rating. This underscores the efficacy of GA-BPNN in effectively classifying the underlying default risk and credit ratings of MSMEs, offering a promising approach for decision-making.

New constructions of decision evaluation functions in three-way decision spaces based on uninorms

In 2014, Hu introduced the concept of three-way decision spaces and axiomatic definition of decision evaluation functions. In three-way decision spaces, decision evaluation function satisfies minimum element axiom, monotonicity axiom and complement axiom. Since then, the research on construction method of decision evaluation functions from commonly used binary aggregation functions becomes a research hotspot. Meanwhile, uninorms, as one class of binary aggregation functions, have been successfully applied in various application problems, such as in decision making, image processing, data mining, etc. This paper continues to consider this research topic and mainly explores the new construction methods of decision evaluation functions based on uninorms. Firstly, we show two novel transformation methods from semi-decision evaluation functions to decision evaluation functions based on uninorms. Secondly, using known semi-decision evaluation functions, we give some new construction methods of semi-decision evaluation functions. Thirdly, we give some novel construction methods of decision evaluation functions and semi-decision evaluation functions related to fuzzy sets, interval-valued fuzzy sets, fuzzy relations and hesitant fuzzy sets. Based on them, decision maker can obtain more useful decision evaluation functions, thereby more choices can be used for realistic decision-making problems. Finally, we consider two real evaluation problems to illustrate the results obtained in this paper. The three-way decisions results of evaluation problem show that the construction method proposed in this paper is superior to some existing construction methods under some conditions.

Assessment of the Exterior Quality of Traditional Residences: A Genetic Algorithm–Backpropagation Approach

The visual aesthetics of villages are remarkably affected by the exterior quality of traditional residences, influencing the impression and assessment of local culture. A proper scientific assessment of exterior quality can protect traditional cultures and improve the development of villages. This research was conducted in a village consisting of 115 residences (Mengjinglai village, which is on the border between China and Myanmar). The backpropagation (BP) neural network model with genetic algorithm (GA) was applied to evaluate the quality of the dwellings. All the evaluation values of the dwellings were defined by scores. Meanwhile, the score of each residence was affected by three main factors: architectural spatial elements, architectural construction elements, and historical and cultural elements. The results show that the village’s dwellings are well preserved and clearly express the traditional Dai style. Moreover, the GA–BP approach is more suitable than the traditional BP method for the assessment of the exterior quality. The quantitative machine learning model would be useful for other aspects of the assessment of similar villages in the future.

Design of an Intelligent Nursing Information Management System for Critically Ill Patients in Neurosurgery

In hospitals, one of the dominant issues is the development of an accurate and precise nursing management system which is hard to implement due to the various problems in the implementation of the traditional manual system. For this purpose, we are to solve the imperfect functions of the traditional nursing information management system and the strong subjectivity and low accuracy of the way of manually judging the patient's condition. Firstly, the Immune Genetic Algorithm (IGA) is used to optimize the Backpropagation Neural Network (BPNN). A mortality prediction model using IGA-BPNN is proposed. Secondly, a nursing information management system for critically ill patients in neurosurgery is designed. The IGA-BPNN prediction model is used as a part of the system to predict the mortality of critically ill patients. Finally, the performance of the predictive model and the system is tested using the Medical Information Mart for Intensive Care (MIMIC)-III data set design experiment. The results show the following: (1) Precision, Recall, and F1-score of mortality prediction using the IGA-BPNN model are 7.2%, 7.2%, and 7.3% higher than those of other prediction models. The designed model has better performance. (2) The comprehensive performance of the system during operation can reach the standard. The researched content aims to provide important technical support for the nursing information management of critically ill patients in neurosurgery and the intelligent analysis of patients' condition.

RODFormer: High-Precision Design for Rotating Object Detection with Transformers

Aiming at the problem of Transformers lack of local spatial receptive field and discontinuous boundary loss in rotating object detection, in this paper, we propose a Transformer-based high-precision rotating object detection model (RODFormer). Firstly, RODFormer uses a structured transformer architecture to collect feature information of different resolutions to improve the collection range of feature information. Secondly, a new feed-forward network (spatial-FFN) is constructed. Spatial-FFN fuses the local spatial features of 3 × 3 depthwise separable convolutions with the global channel features of multilayer perceptron (MLP) to solve the deficiencies of FFN in local spatial modeling. Finally, based on the space-FFN architecture, a detection head is built using the CIOU-smooth L1 loss function and only returns to the horizontal frame when the rotating frame is close to the horizontal, so as to alleviate the loss discontinuity of the rotating frame. Ablation experiments of RODFormer on the DOTA dataset show that the Transformer-structured module, the spatial-FFN module and the CIOU-smooth L1 loss function module are all effective in improving the detection accuracy of RODFormer. Compared with 12 rotating object detection models on the DOTA dataset, RODFormer has the highest average detection accuracy (up to 75.60%), that is, RODFormer is more competitive in rotating object detection accuracy.

NetNMSP: Nonoverlapping maximal sequential pattern mining

Nonoverlapping sequential pattern mining, as a kind of repetitive sequential pattern mining with gap constraints, can find more valuable patterns. Traditional algorithms focused on finding all frequent patterns and found lots of redundant short patterns. However, it not only reduces the mining efficiency, but also increases the difficulty in obtaining the demand information. To reduce the frequent patterns and retain its expression ability, this paper focuses on the Nonoverlapping Maximal Sequential Pattern (NMSP) mining which refers to finding frequent patterns whose super-patterns are infrequent. In this paper, we propose an effective mining algorithm, Nettree for NMSP mining (NetNMSP), which has three key steps: calculating the support, generating the candidate patterns, and determining NMSPs. To efficiently calculate the support, NetNMSP employs the backtracking strategy to obtain a nonoverlapping occurrence from the leftmost leaf to its root with the leftmost parent node method in a Nettree. To reduce the candidate patterns, NetNMSP generates candidate patterns by the pattern join strategy. Furthermore, to determine NMSPs, NetNMSP adopts the screening method. Experiments on biological sequence datasets verify that not only does NetNMSP outperform the state-of-the-arts algorithms, but also NMSP mining has better compression performance than closed pattern mining. On sales datasets, we validate that our algorithm guarantees the best scalability on large scale datasets. Moreover, we mine NMSPs and frequent patterns in SARS-CoV-1, SARS-CoV-2 and MERS-CoV. The results show that the three viruses are similar in the short patterns but different in the long patterns. More importantly, NMSP mining is easier to find the differences between the virus sequences.