Current Issue

All articles published in this issue have undergone a thorough peer review process, and stringent checks for repetition rates have been implemented to ensure the integrity of the content.

Total number of articles in this issue: 7
Total number of pages in this issue: 49

For inquiries regarding the content of specific articles, please feel free to contact the respective authors via their provided email addresses. For questions related to the journal itself, please reach out directly to SUAS Press.

Year

2024

Volume

2

Number

6

Status

Archived

Published

2024 November 16

Articles

Open Access Research ArticlePublished 16 November 2024

Optimizing Urban Road Networks for Resilience Using Genetic Algorithms

Xueyi Cheng, Chang Che

Urban road networks face increasing challenges in balancing traffic efficiency, budget limitations, and environmental impacts as cities prepare for future demand. This paper presents a multi-objective optimization approach using Genetic Algorithms (GAs) to enhance the performance of an urban transportation network while integrating sustainability goals. By simultaneously optimizing travel times, reducing bottlenecks, and addressing budget constraints, this framework enables a balanced approach to infrastructure improvement. The inclusion of environmental considerations, such as greenhouse gas (GHG) emissions, aligns network development with broader sustainability objectives, promoting a healthier urban environment. Future extensions of this framework include adaptive strategies to respond to shifting traffic patterns and the potential integration of regulatory constraints, such as emission licenses. The proposed GA approach demonstrates a flexible, scalable solution for urban planners and policymakers tasked with building resilient, sustainable road networks, offering practical insights into addressing the multifaceted demands of modern urban infrastructure.

Disciplines:ManagementSubjects:Management OptimizationReferences:31

DOI10.5281/zenodo.14032011ARKark:/40704/AJSM.v2n6a01
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

Machine Learning-Driven Fraud Detection: Management, Compliance, and Integration

Xueyi Cheng

This research delves into the comprehensive methodology of employing machine learning in the domain of fraud detection, outlining the critical steps from data collection to continuous learning. It emphasizes the importance of adhering to data protection regulations during the data collection phase and the significance of preprocessing in preparing the data for analysis. The study explores various machine learning models, including supervised and unsupervised learning techniques, and evaluates their performance using metrics such as accuracy and AUC-ROC. It highlights the necessity of continuous learning to adapt to evolving fraud tactics and the challenges of integrating machine learning models into existing fraud detection systems. Ultimately, this research underscores the transformative potential of machine learning in enhancing the accuracy and efficiency of fraud detection, safeguarding financial transactions, and protecting consumers from fraudulent activities.

Disciplines:EconomicsSubjects:Financial Risk ManagementReferences:20

DOI10.5281/zenodo.14064121ARKark:/40704/AJSM.v2n6a02
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

Enhancing Stock Price Prediction through Attention-BiLSTM and Investor Sentiment Analysis

Kangming Xu, Biswajit Purkayastha

The change of stock price is the focus of investors in the stock market, so stock price trend prediction has always been a hot topic in quantitative investment research. Traditional machine learning prediction model is difficult to deal with nonlinear, high frequency and high noise stock price time series, which makes the prediction accuracy of stock price trend low. In order to improve the forecasting accuracy, the temporal characteristics of stock price data are studied. A bidirectional long short-term memory neural network combining empirical mode decomposition (EMD), investor sentiment and attention mechanism is proposed to predict the rise and fall of stock prices. First, the empirical mode decomposition algorithm is used to extract the characteristics of stock price time series on different time scales, and the investor complex index of the text from the close of the last trading day to the opening of the next trading day is extracted by constructing the all-inclusive sentiment dictionary.The realization of a stock price trend prediction model based on Attention-BiLSTM involves combining the Bidirectional Long Short-Term Memory (BiLSTM) network with an attention mechanism. The BiLSTM processes data points from both past and future for better context understanding, while the attention mechanism selectively focuses on crucial information, improving the model's predictive accuracy in capturing and utilizing patterns in stock price movements. This sophisticated approach enhances the model's ability to forecast stock trends effectively.

Disciplines:FinanceSubjects:Financial ManagementReferences:15

DOI10.5281/zenodo.14065931ARKark:/40704/AJSM.v2n6a03
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

Integrating Artificial Intelligence with KMV Models for Comprehensive Credit Risk Assessment

Kangming Xu, Biswajit Purkayastha

With the continuous development of artificial intelligence and various new intelligent algorithm technologies, the business contacts between various institutions within financial enterprises are gradually increasing, and traditional financial risk management can no longer adapt to the current status quo in the era of big data. The lack of information sharing among institutions can reduce the efficiency of financial management and adversely affect the operation of enterprises. At present, financial credit risk mainly includes credit risk, market risk and operational risk. Credit risk relates to the possibility that a borrower will not be able to repay loans or debts on time, market risk covers potential losses caused by market volatility, price changes and adverse events, while operational risk includes risks such as internal operational errors, technical failures and fraud, which may adversely affect the normal operations and financial condition of a financial institution. These risk factors need to be integrated and managed in the financial sector to ensure financial stability and customer trust. Therefore, this paper aims to establish a KMV financial credit risk model, continuously strengthen the internal risk management of enterprises, achieve management modeling and a good KMV algorithm mechanism, and realize the cooperation and stickiness between customers and enterprises, so as to avoid unnecessary financial risks.

Disciplines:FinanceSubjects:Risk ManagementReferences:20

DOI10.5281/zenodo.14077150ARKark:/40704/AJSM.v2n6a04
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

A Graph Neural Network-Based Approach for Detecting Fraudulent Small-Value High-Frequency Accounting Transactions

Yitian Zhang, Yibang Liu, Shuaiqi Zheng

The growth of digital accounting systems has led to increased fraud schemes, especially those involving small-value businesses. This paper presents a novel neural network architecture to capture order to break it down to break up suitable class and image heterophily in fraud detection by holding different representations for homophilic and heterophilic characteristics, making it more effective in detecting fraud patterns. The model includes a unique system of body-aware construction and adaptive memory to capture complex changes on multiple time scales. We introduce a two-channel feature extraction mechanism that performs similar and different processes independently, facilitating the storage and propagation of fraud signals from the business network. Various experiments on two real-world datasets show that our method significantly improved over the state-of-the-art method, with a performance of 12.3% in AUC -ROC and 15.7% in F1-score. The model is particularly effective in identifying fraud schemes that use multiple accounts and different currencies, achieving a 67% reduction in false positives. Our results show the model can identify subtle transaction patterns that distinguish fraudulent from legitimate transactions.

Disciplines:FinanceSubjects:Financial Risk ManagementReferences:35

DOI10.5281/zenodo.14161459ARKark:/40704/AJSM.v2n6a05
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

Research on Deep Learning-Based Authentication Methods for E-Signature Verification in Financial Documents

Yining Zhang, Wenyu Bi, Runze Song

This paper presents a novel deep learning-based authentication method for electronic signature verification in financial documents. The proposed system introduces a comprehensive framework integrating YOLOv5-based signature detection, advanced preprocessing techniques, and a Siamese neural network architecture for verification. The system employs a customized feature extraction network incorporating residual connections and attention mechanisms to capture local and global signature characteristics. The implementation includes adaptive preprocessing pipelines and sophisticated loss functions optimized for signature verification tasks. Experimental evaluation on a dataset of 25,000 signature samples from 500 individuals demonstrates superior performance, achieving 98.5% accuracy in verification tasks with a false acceptance rate of 1.2% and a false rejection rate of 1.5%. The system maintains robust performance across various document conditions, demonstrating only a 4.2% accuracy reduction under poor resolution scenarios. Security analysis validates system resilience against adversarial attacks, achieving a 96.5% detection rate. The comprehensive evaluation demonstrates significant improvements over existing accuracy and computational efficiency methods, establishing new benchmarks for signature verification in financial applications. The proposed methodology addresses critical challenges in financial document security while maintaining practical applicability in real-world environments.

Disciplines:FinanceSubjects:Financial Risk ManagementReferences:33

DOI10.5281/zenodo.14161744ARKark:/40704/AJSM.v2n6a06
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 16 November 2024

The Core Secrets of Crafting the "Dramatic Poetic Realm" in The Romance of the Western Chamber

Mingze Lu

This article takes The Romance of the Western Chamber as the research object, and makes a specific analysis on how to construct its dramatic poetic environment. The methods of creating drama are embodied in two aspects: "the ups and downs and the differences are interesting", "the potential energy is full of drama flavor", and the methods of creating poetry are embodied in the two aspects of "the combination of poetry and painting" and "the fullness of image adds spiritual rhyme", which form a dramatic poetic environment. At the end of the article, combining with many problems existing in the current drama creation, the author puts forward several constructive suggestions from the perspective of constructing the drama poetic environment in The Romance of the Western Chamber.

Disciplines:HumanitiesSubjects:Philosophy of ArtReferences:7

DOI10.5281/zenodo.14197079ARKark:/40704/AJSM.v2n6a07
View ArticleDownload PDFScholar

Copyright © 2026 The Author(s). Published by Southern United Academy of Sciences.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Academic Journal of Sociology and Management (AJSM)
Published by Southern United Academy of Sciences Limited ISNI: 0000000512776460, operated by the Publications Division.