International Journal of Innovative Technology and Interdisciplinary Sciences

Deep Excavation Protection for Urban Underground Construction: A Case Study of an Anchored Secant Pile Wall

Mon, 03 Aug 2026 00:00:00 +0200

Modern urban underground construction requires retaining systems that satisfy both ultimate and serviceability limit states while ensuring reliable performance throughout excavation. This study presents the numerical evaluation and performance verification of an anchored secant pile wall through the case study of the deep excavation at “Çerçiz Topulli” Square, Gjirokastra, Albania. The study involved a 12.9 m deep excavation for a three-level underground structure in multilayered soil conditions. A reproducible plane-strain finite-element model, incorporating beam-on-elastic-foundation elements, was developed to simulate staged excavation, soil stratification, groundwater conditions, anchor activation, and construction sequences. The retaining system performance was assessed using normalized wall deflection, serviceability criteria, deformation prediction accuracy, and anchor efficiency. The maximum calculated lateral wall displacement was 12.1 mm, corresponding to a normalized wall deflection (δmax/H) of 0.094%, which is substantially below the commonly accepted serviceability limit of 0.5%. Comparative analyses indicate that the prestressed anchor system reduced wall deformation by approximately 25% relative to an equivalent unanchored excavation, while controlled dewatering was required to maintain excavation stability. The proposed framework provides a validated methodology for the design verification and performance assessment of anchored secant pile walls in deep urban excavations. The findings offer practical guidance for optimizing retaining systems in multilayered soils and establish a transferable approach for similar underground infrastructure projects in complex urban environments.

An Empirical Decision-Support Framework Integrating Statistical Learning and Digital Twin for Predictive Maintenance Planning in Industry 4.0

Ram Ballabh Sinha, Vimal Bhatt, Satya Nand Jha — Sat, 08 Aug 2026 00:00:00 +0200

The integration of Digital Twin technology with advanced analytics has created new opportunities for predictive maintenance in Industry 4.0. This study develops and empirically validates a framework that examines how statistical learning capability and Digital Twin integration influence predictive maintenance performance. Data were collected through a quantitative cross-sectional survey of 210 industry professionals, and the proposed relationships were evaluated using Structural Equation Modelling (SEM) with bootstrapping. The findings show that statistical learning capability, when integrated, greatly improves predictive maintenance efficiency (β = 0.61, p < 0.001). Additionally, incorporating Digital Twin into predictive maintenance affects both predictive efficiency (β = 0.47, p < 0.001) and fault detection accuracy (β = 0.58, p < 0.001). Maintenance performance is largely mediated by the influence of predictive maintenance efficiency (β = 0.66, p < 0.001) on the link between statistical learning capability and performance (indirect effect β = 0.40, p < 0.001). The model explains well and adds to the body of literature providing an empirically proven relationship between statistical learning and Digital Twin technologies. The cross-sectional design of the study, however, may limit the generalizability of the results and studies based on real-time data and longitudinal data across sectors are recommended. In a practical sense, the results demonstrate the importance of having an integrated analytical and digital strategy for organizations to improve predictive capability, minimize downtime, and optimize maintenance planning.

Comparative Techno-Economic Feasibility Assessment of Onshore Wind Energy Deployment: A Multi-Site Analysis

Elena Bebi, Majlinda Alcani, Ardit Gjeta, Altin Dorri, Tauland Spahiu — Sat, 08 Aug 2026 00:00:00 +0200

Given Albania’s high dependence on hydropower and its sensitivity to hydrological fluctuations, identifying alternative renewable energy sources has become increasingly important. This study presents a comparative techno-economic feasibility assessment of onshore wind farm development in ten selected areas in Albania. Since studies simultaneously evaluating the technical and economic feasibility of multiple potential wind energy sites remain limited in Albania, this work provides a broader assessment of wind energy development feasibility. Wind resource assessment was performed using long-term wind atlas-based data where available complemented by on site measurements. WAsP software was used to assess the wind energy resources and to simulate the wind farm energy output. Weibull methodology was used for probabilistic assessment of wind speeds. Five wind turbine models were simulated. The Vestas V163 4.5 MW turbine at 126 m hub demonstrated the best technological and economic overall performance and was selected as the reference wind turbine for the economic assessment. The economic assessment was performed using RETScreen Expert software while sensitivity analyses were performed with respect to variations in installation cost, electricity price and discount rate. Wide range of values of the technical wind farm indicators and economic indicators was reported across the studied sites. Vau i Dejës, Mali i Rencit and Pukë can be considered to have the most favourable technical and economic indicators due to their high wind farm capacity factor and low LCoE value, combined with high NPV and IRR values. The LCoE varied from approximately 49 to 63 €/MWh. Notably, Mali i Rencit emerged as one of the best-performing sites despite not belonging to the highest wind power classes, indicating that feasibility also depends on other site-specific factors beyond wind potential. Furthermore, the comparison between WAsP and RETScreen highlighted the importance of a detailed and representative wind resource assessment for improving the reliability of wind farm evaluations. Overall, the findings provide a useful basis for identifying priority areas and supporting future investment decisions in Albania’s wind energy sector.

Public Debt and Foreign Direct Investment in Transition Economies: A Data-Driven Comparative Analysis of Crisis Dynamics and Threshold Effects

Bardhyl Dauti, Drita Krasniqi, Olcay Çolak, Rametulla Ferati — Sun, 09 Aug 2026 00:00:00 +0200

This study examines the relationship between public debt and foreign direct investment (FDI) in six Western Balkan (WB-6) countries and eleven Central and Eastern European Union (CE-EU-11) countries during the period 2000–2023. The results, based on a dynamic panel threshold regression model with endogenous regressors, indicate that public debt has a statistically significant adverse effect on FDI across both debt regimes, with the estimated threshold ranging between approximately 69% and 72% of GDP. These findings provide evidence of a persistent crowding-out effect of public debt on foreign investment. While the magnitude of the effect varies across debt regimes, public debt consistently discourages FDI inflows. The findings also reveal strong persistence in FDI, with moderate inflation and institutional quality exerting positive effects. Furthermore, the positive interaction between EU membership and public debt suggests that foreign investors are more tolerant of higher public debt levels in Central and Eastern European EU countries than in the Western Balkans.

An Efficient Method for Lung Cancer Classification Using EfficientNet and Learner Cooking Optimization

Chakka Srividya, Krishnamoorthy Ramasubramanian, Myneni Madhu Bala — Fri, 14 Aug 2026 00:00:00 +0200

Computational approaches are increasingly important for accurate and privacy-preserving analysis of medical imaging data. This study proposes a federated deep learning framework that integrates optimized deep feature extraction, feature selection, and privacy-preserving collaborative learning for multiclass lung cancer classification from CT images. The proposed method uses convolutional neural network with compound scaling by performing hierarchical feature extraction and then feature selection strategy to minimize redundancy and improve the class separability is based on optimization. The training of the model occurs in a federated (between several clinical nodes) distributed environment in which a global classifier is jointly trained (learned) without any actual exchange of patient images, a factor that guarantees the data privacy. Experiments with the publicly available LIDC-IDRI prove that the proposed framework has a total accuracy of 99.73, precision of 99.62, sensitivity of 99.82, specificity of 99.43, and F1-score of 99.38 with four lung cancer types. The architecture of the proposed federated feature aggregation framework is not specific to lung CT imaging, the pattern of combining a pretrained deep feature extractor, a bio-inspired feature optimizer, and secure additive masking is structurally applicable to other privacy-sensitive medical imaging tasks, though validation in those domains would require independent experimental evidence and is identified as future work.

From Prediction to Action: Explainable Churn Modelling with Calibration and Threshold-Aware Decision Support

Frida Zisko, Erarda Vuka — Tue, 18 Aug 2026 00:00:00 +0200

The problem of predicting customer churn plays an important role in the field of telecommunication business decisions. However, the existing literature mainly focuses on achieving better classification accuracy without paying enough attention to calibration, interpretability, and the business value of threshold choice. The goal of this study is to introduce a churn prediction framework that incorporates optimization of the classifier, its calibration, explainability, and business-oriented choice of the threshold. The open-source Telco Customer Churn dataset was used to compare four models like Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM). The process of hyperparameter tuning is carried out by means of randomized search and stratified cross-validation. Further experiments include repeated stratified cross-validation and hold-out validation of the models' results. Random Forest showed the best performance in terms of mean cross-validation, while LightGBM demonstrated the best F1-score and Matthew’s correlation coefficient on the hold-out set. Due to the lack of statistical significance of the difference between these algorithms, LightGBM was chosen for further use. Isotonic calibration decreases the Brier score from 0.1609 to 0.1371. The threshold optimization shows that choosing the decision boundary equal to 0.30 improves the balance between precision and recall compared to the usual 0.50. Furthermore, the cost-sensitive analysis demonstrates that the optimal value of thresholds decreases as the cost of missing out on a true churner rises. SHapley Additive exPlanations (SHAP) values and permutation-based feature importance were calculated to identify the four most important predictors of churn, including contract type, customer tenure, internet service type, and billing variables. In this research, a decision support framework is developed by calibrating probabilities and tuning decision thresholds.