Advancement in Customer Attrition Prediction: Design of Optimal Triple Refined Indeterminate Neutrosophic Sets in Large-Scale Financial Sectors

Bunyodbek; Dilrabo; Hamdam; Natalia; K.

doi:https://doi.org/10.54216/IJNS.260204

Journal of Intelligent Systems and Internet of Things | 2025

Leveraging Quantum Neural Networks with Deep Learning Based Edge Detection Model for Breast Cancer Screening using Digital Mammograms

Breast cancer (BC) is one of the most common invasive cancers, which cause thousands of women's deaths globally. Therefore, prompt detection is a cure for reducing the rate of death. Therefore, screening of BC in its initial phase is of utmost vital. Physically segmenting breast lesion imaging appears a time-consuming and expensive pursuit for radiologists. Hence, the adoption of automatic analytic techniques becomes vital, directing to exactly segment lesions of the breast and mitigate the associated tasks. The segmentation of malignant areas is an essential procedure in the complete inspection of breast image data. To achieve the segmentation and recognition of BC, numerous computer-aided diagnosis (CAD) techniques were presented for the investigation of mammogram imaging. The CAD models are employed to mainly analyze the disorder and provide the best treatment. Currently, deep learning (DL) techniques are superior and provide promising results in the early recognition of BC. In this paper, we design a Leveraging Quantum Algorithms for Edge Detection in Mammograms to Improve Breast Cancer Screening (LQAEDM-IBCS) model. The main intention of the LQAEDM-IBCS is to provide an accurate and effective technique for the detection and segmentation process of breast cancer using advanced algorithms. Initially, the image pre-processing stage applies the adaptive bilateral filtering (ABF) method to eliminate the unwanted noise in input image data.  Next, the segmentation process is implemented by the Otsu threshold method for edge detection. To improve the segmentation performance, the parameter tuning process is performed through the quantum spotted hyena optimizer (QSHO) algorithm. Besides, the proposed LQAEDM-IBCS technique designs the DenseNet-121 method for the extraction of feature procedure. Eventually, the quantum neural network (QNN) method has been deployed for the BC classification process. The simulating validation of the LQAEDM-IBCS system is verified on a benchmark image database and the outcomes are dignified under numerous measures. The experimental outcome emphasized the enlargement of the LQAEDM-IBCS approach in the BC diagnosis process.

groups

S. Abdel-Khalek mail

link https://doi.org/10.54216/JISIoT.160206

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Real-Time Classroom Emotion Analysis Using Machine and Deep Learning for Enhanced Student Learning

    This research creates an innovative EfficientNet-B7-based Facial Expression Recognition model that delivers maximum accuracy performance for detecting emotions. Successful classification performance benefits substantially from EfficientNet-B7's application of compound scaling techniques which balances the entire network dimensions depth width and resolution. The characteristic distinctive to EfficientNet-B7 over standard architectural models involves its dual capability to perform accurate computations at reduced complexity levels. The model receives evaluation using KDEF at high-resolution as well as FER2013 at low-resolution through usage of SGD, Adam, and RMSprop optimizers. Experimental tests confirmed that EfficientNet-B7 operates with RMSprop optimizer to recognize emotions on KDEF at 91.78% accuracy superior to ResNet152's highest recorded accuracy of 88.77%. Performance levels declined to 57.56% on FER2013 because low-resolution images represent a great challenge to the model. Internal Batch Normalization (IBN) enters the model as an issue solution to halt gradient descent problems, which results in better model training stability and enhanced accuracy-loss patterns. The research demonstrates that FER performance benefits greatly when EfficientNet-B7 works in combination with IBN for high-resolution image processing. The research proves that EfficientNet-B7 stands as a reliable FER solution that shows potential usage in affective computing and human-computer interaction domain.  

groups

Deepa devasenapathy mail -

Krishna Bhimaavarapu mail -

Prem Kumar Sholapurapu mail -

S. Sarupriya mail

link https://doi.org/10.54216/JISIoT.160207

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Research on the Evaluation Method of Energy Sustainable Development Indicator System Based on Genetic Algorithm and Local Support Vector Regression

With the acceleration of modern urbanisation, the demand for energy by the state and society is increasing. In order to maintain the sustainable availability of energy, it is necessary to establish an energy sustainability indicator system. To address this issue, this paper proposes an innovative evaluation method for energy sustainability indicator system, which aims to provide a multi-scale and more comprehensive assessment of energy sustainability indicators, as well as to ensure the accuracy and reliability of the evaluation results. This paper proposes to use genetic algorithm and local support vector regression (SGA-LSVM) to optimise the projective fuzzy clustering solution model (PPFCM), to establish a new evaluation method of energy sustainability index system based on genetic algorithm and local support vector regression. Based on this method, energy sustainability in different regions is analyzed according to three indicators: energy supply side, demand side and affordability, and the validity of this evaluation method is tested. The study found that, in terms of zoning: the eastern region is in the lead in energy demand side, energy supply side and energy affordability, and the western region has a rising trend in recent years; in terms of population density: the indices of energy demand side, energy supply side and energy affordability of densely populated areas are much higher than the rest of the areas compared to the sparsely populated areas, and the difference between the indices of energy supply side and energy affordability of the sparsely populated and moderately populated areas and the difference between the indices is not significant. The energy supply-side index is slightly higher than that of the medium-population area; Economy and carbon emission: due to China's focus on environmental protection, carbon emission is kept within a stable range while the economy is developing rapidly. By PC≥0.80, PE≥0.45 and XB≤0.1, it shows that the method of evaluating the energy sustainable development index system using the fuzzy projection-seeking clustering energy sustainable development evaluation model based on genetic algorithm and local vector regression optimization is reliable.

groups

Qian Chen mail

link https://doi.org/10.54216/JISIoT.160208

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Research on Image Generation Style Transfer and Reconstruction Loss Reduction Based on Deep Learning Framework

Nixi black pottery has a unique place in Chinese black pottery art. In this article, we have developed a style transfer model based on deep learning, which automatically transforms Nixi black pottery into images of other styles. This is of great value for the dissemination of this art. In this paper, we propose a method called DualTrans that utilizes a pure Transformer architecture to enable context-aware image processing, effectively addressing the issue of low receptive field. Additionally, we introduce a Location Information Encoding Module (LIM) and a Style Transfer Control Module (STCM) to tackle the problem of long-range dependencies while ensuring that the generated target image remains structurally and stylistically consistent throughout the style transfer process, without being influenced by the content and style images. During the mapping process, the LIM encodes the original image block information and concatenates it with the projected image block information. To alter the final produced style of the picture, the STCM leverages a set of learnable style-controllable factors. Extensive trials have shown that DualTrans exceeds previous approaches in terms of stability.

groups

Wei Zou mail -

Mohd Alif Ikrami Bin Mutti mail

link https://doi.org/10.54216/JISIoT.160209

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Comment Feedback Optimization Algorithm (CFOA): A Feedback-Driven Framework for Robust and Adaptive Optimization

The Comment Feedback Optimization Algorithm (CFOA) presented a novel feedback-driven model for solving optimization problems, incorporating ideas based on positive and negative feedback loops. Unlike other optimization algorithms, CFOA includes feedback adjustments for better tuning the exploration-exploitation trade-off, thus making CFOA less sensitive to the dimensions of problems and their nonlinearity. Some proposed features include feedback dynamics for adaptive search options, parameter control by a decay function, and mechanisms for escaping local optima. CFOA’s performance has been benchmarked on CEC 2005 test cases with many evaluations. The results demonstrate better convergence speed, solution quality, and computational complexity compared with the Sine Cosine Algorithm (SCA), Gravitational Search Algorithm (GSA), and Tunicate Swarm Algorithm (TSH). The efficiency of the approach used by CFOA makes it an indispensable tool for solving real-world optimization problems across various application domains such as machine learning, engineering, and logistics.

groups

El-Sayed M. El-kenawy mail -

Amel Ali Alhussan mail -

Doaa Sami Khafaga mail -

Amal H. Alharbi mail -

Sarah A. Alzakari mail -

Abdelaziz A. Abdelhamid mail -

Abdelhameed Ibrahim mail -

Marwa M. Eid mail

link https://doi.org/10.54216/JISIoT.160210

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

Journal of Intelligent Systems and Internet of Things | 2025

Greylag Goose Optimization for Diabetes Prediction: Feature Selection Meets Advanced Machine Learning

Diabetes mellitus remains a global health concern, necessitating both accurate and effective diagnostic methodologies. This condition presents a significant challenge due to the high dimensionality of clinical datasets and the inherent complexity of diabetes classification. To address this problem, this study integrates feature selection and machine learning architectures to enhance diabetes prediction accuracy. A novel framework based on the Binary Greylag Goose Optimization (bGGO) algorithm is proposed to optimize feature selection, thereby improving classification performance. A comprehensive evaluation uses multiple classifiers, including Decision Trees, k-nearest Neighbors, Support Vector Machines, Random Forests, and Multilayer Perceptron (MLP). The experimental results demonstrate that bGGO significantly enhances feature selection quality, improving classification metrics, particularly for MLP, which achieves the highest classification accuracy of 95.98%. These findings underscore the efficacy of combining metaheuristic optimization with machine learning for diabetes diagnosis, offering a scalable and interpretable approach for real-world healthcare applications. The proposed methodology contributes to more precise risk estimation and the development of individualized intervention strategies, facilitating early diagnosis and effective disease management.

groups

Gomaa Mohamed Ismail mail -

El-Sayed M. El-Kenawy mail -

Shady Y. El-Mashad mail

link https://doi.org/10.54216/JISIoT.160202

Volume & Issue

Vol. Volume 16 / Iss. Issue 2

Details open_in_new

International Journal of Neutrosophic Science | 2025

Differential Equation of COVID-19 with Constraint Algebraic Equation and Sustainable health development With Applications in Neutrosophic Environment

The first appearance of COVID-19 in late 2019 and spread rapidly throughout the world until it became a global pandemic, and the World Health Organization announced some vaccines, and the emergence of a mutated version of COVID-19 was reported in several countries, including Iraq, and we will take care of conducting a study on the spread and dynamics of a virus, this work will be based on the study of the dynamics 3D harvesting predator (COVID-19) differential-algebraic predator-prey economic model (DA-PPM) with functional responses of Holing type-II. The appropriate and realistic description with high accuracy of this phenomenon, which may be natural and emerging as such models, has proven the sentimentality and existence of the solution to the system, and the stability of the system, was discussed in a manner similar to the stability of Matignon. The numerical results showed that the variables of stable unhappy situations have an effect, and this important study can be used as one of the methods of health science to control the spread of COVID-19 and its advanced models.  One of the critical aspects of sustainable development is building resilient health systems capable of dealing with epidemics and other crises, the mathematical model (DA-PPM) was applied to analyze the sustainability of health systems under the pressure of Covid-19 and evaluate how long-term public health policies and interventions can prevent overexploitation of resources. Ensuring equitable access to care. The application of the mathematical model to understand the spread of the epidemic is discussed to observe the spread of the epidemic, the possibility of coexistence with it, its close relationship with sustainable development, and to emphasize the importance of the flexibility of the health system. In addition, we apply our results on the neutrosophic supposed data that deals with uncertainty in real-life measurements and compare it with the classical results.

groups

Mohammed Kadhim Mohsin mail -

A. Y. J. Almasoodi mail -

Sarah A.AL-Ameedee mail -

Mohammed Qassim mail

link https://doi.org/10.54216/IJNS.260201

Volume & Issue

Vol. Volume 26 / Iss. Issue 2

Details open_in_new

International Journal of Neutrosophic Science | 2025

New Class of Equivalence Classes of Neutrosophic Fuzzy Delta- Algebras

This work analyzes neutrosophic fuzzification in algebra, applies novel classes of neutrosophic fuzzy () to algebra, and explores the ideas of ideal (), subalgebra (), δ-homomorphism, and ideal (), exploring some of their descriptions. We shall demonstrate a variety of applications, including the notations  and  on  /  is  of }. We will also investigate their equivalence classes, evaluate our findings in light of the unique ideas offered in this work, and investigate related characteristics.

groups

Azeez Lafta Jaber mail -

Hussein S ALallak mail -

Jaafer Hmood Eidi mail -

Shuker Khalil mail

link https://doi.org/10.54216/IJNS.260202

Volume & Issue

Vol. Volume 26 / Iss. Issue 2

Details open_in_new

International Journal of Neutrosophic Science | 2025

Correlation Measure for NeutroSophic Filter in Medical Diagnosis

The aim of this article is proposed the notion of Correlation Measure for Neutrosophic Filter (NF). Additionally using correlation measure for neutrosophic set the application of medical diagnosis were discussed with numerical example.

groups

P. Susithabanu mail -

V. Nirmala mail

link https://doi.org/10.54216/IJNS.260203

Volume & Issue

Vol. Volume 26 / Iss. Issue 2

Details open_in_new

International Journal of Neutrosophic Science | 2025

Advancement in Customer Attrition Prediction: Design of Optimal Triple Refined Indeterminate Neutrosophic Sets in Large-Scale Financial Sectors

Background: Neutrosophy is the subject area of philosophy that researches all associated with neutralities, owing to the contradictory information, lack of information, imprecise and paradoxical information, among them. The scale's design is organized to take the subjective quality of opinion, being responsible for either uncertainty or the indeterminacy of the respondents' opinions. It relies on the triple refined indeterminate neutrosophic sets (NS) for improved accuracy in understanding the agreement or disagreement level on particular items, like the competence of activities cost and financial management inside the legal services. Currently, customer abrasion is more and more serious in commercial banks, mainly, high-valued customers in retail banking. Therefore, it is stimulated to advance a prediction mechanism and recognize this customer may be at attrition risk. Thus, recognizing and lowering customer churn has become important for financial institutions trying to maintain customers. Currently, several researchers concentrate on customer attrition rate studies utilizing sophisticated machine learning (ML) and deep learning (DL) methods. Methodology: This study develops a Customer Attrition Prediction Using Triple Refined Indeterminate Neutrosophic Sets with an Optimization Algorithm (CAP-TRINSOA) technique. The main aim of the CAP-TRINSOA technique is to improve the attrition prediction of a customer in large-scale financial sectors using state-of-the-art techniques. In the initial stage, the data normalization employs mean normalization to transfer input data into an even format. Furthermore, the classification process is performed by implementing the triple refined indeterminate neutrosophic sets (TRINS). Finally, the honey badger algorithm (HBA) alters the parameter tuning value of the TRINS method optimally and results in greater performance of classification. Results: An extensive set of simulations is accomplished to exhibit the promising results of the CAP-TRINSOA method under the bank customer churn prediction dataset. The experimental validation of the CAP-TRINSOA technique portrayed a superior accuracy value of 97.65% over exisitng model in the customer attrition prediction process.

groups

Bunyodbek Sultonov mail -

Dilrabo Akhmedova mail -

Hamdam Matyaqubov mail -

Natalia Falina mail -

K. Shankar mail

link https://doi.org/10.54216/IJNS.260204

Volume & Issue

Vol. Volume 26 / Iss. Issue 2

Details open_in_new