Wireless sensor network (WSN) is an advanced technology in the current scenario owing to its broad range of research. Due to constraints such as restricted bandwidth and ever-changing network structures, WSNs are inherently exposed to a wide range of security vulnerabilities. This inherent fragility has sparked a significant surge in research efforts focused on enhancing the security mechanisms of WSNs in recent years. Thus, the research on WSN security has been growing for the past few years. In terms of security, the less infrastructure and self-reliant nature of WSN is considered a difficult concern. A wormhole (WH) attack detection system on Networked Control Systems (NCSs) is developed to conquer this issue by employing the Secretary Pufferfish Optimization Algorithm enabled Dense ResNeXt fused Deep Stacked Autoencoder (SPOA_DResNeXt-DSAE). Firstly, WSN simulation is performed and routing is executed using Low Energy Adaptive Clustering Hierarchy (LEACH). In order to perform WH attack detection, three processes, like Neighbour Ratio Threshold (NRT), out-of-band and in-band WH detection, are conducted. In the final phase, detection of the WH attack is effectively carried out through the application of the ResNeXt-DSAE framework. Additionally, the attack mitigation is done by means of DResNeXt-DSAE, which is trained using the Secretary-Pufferfish Optimisation Algorithm (SPOA). The effectiveness of DResNeXt-DSAE is evaluated using throughput, delay and Packet Delivery Ratio (PDR), which observed better values of 0.570 sec, 0.704 Mbps and 0.882.

The growing demand for sustainable, pure-cotton products in offline retail necessitates accurate demand forecasting, yet the complex interplay of weather, holidays, and discounts, and their nonlinear and heterogeneous effects, remains underexplored. This study addresses this gap by quantifying multi-factor impacts and developing a two-stage framework integrating empirical analysis with machine learning to leverage nonlinear effects of external factors. We first conduct hierarchical regression using authentic sales data from a leading cotton textile company to identify statistically significant features and regional heterogeneity, which reveals that features such as the quadratic effects of temperature are highly significant, with some product categories sensitive to specific weather ranges. Guided by empirical results, a hybrid model integrates feature engineering and regional segmentation to capture heterogeneities using ensemble learning. This integration of evidence-based variable engineering allows our model to capture complex interactions, adapt to new products and regional variability, and achieve an average accuracy (ACU) of 0.72, a 73% improvement over the moving average baseline. These actionable insights contribute to sales strategy in offline retail and demonstrate the power of making interpretable predictive decisions grounded in robust empirical analysis.

The performance-based warranty can not only protect customers against uncertainty regarding product performance but also assist manufacturers in establishing credibility and promoting sales. This study first develops a novel performance-based warranty policy that integrates limits on the number of repairs for both performance failure and customer-induced accidental failure, aiming to safeguard customer rights and mitigate potential risks for the manufacturer. When the total number of repair actions for performance failures exceeds the threshold, we offer customers flexible compensation options. We use the expected total maintenance cost during the cycle to determine the optimal warranty length and guaranteed performance level from the manufacturer’s perspective. Through numerical experiments, we have derived several interesting practical implications: (i) The manufacturer’s top priority is to continuously improve product performance to reduce the potential number of performance failures. (ii) When limits on the number of repairs for performance failures are lower, the manufacturer is advised to provide more reliability information to help customers evaluate repair expenses more accurately. This mutual exchange benefits both parties by reducing the expected cost during the maintenance cycle.

Charging consumers reasonable prices and compensating service providers fairly are crucial for the successful operation of on-demand service platforms. These factors are also central to a platform’s ability to generate substantial profits. Information plays a pivotal role in platform decision-making. However, due to the complex dynamics of both the consumer and supply markets, along with the fluctuating behaviors of consumers and service providers, and cost considerations, it is challenging for platforms to fully characterize demand or supply. Nevertheless, platforms can relatively easily and accurately obtain partial information about the distribution functions, such as the upper and lower bounds, and the mean. Assuming the demand or supply function follows a multiplicative structure and only partial information (upper and lower bounds, and the mean) is available, we apply a distribution-robust optimization approach to model platform decision-making under two perspectives: pessimistic (maximin) and optimistic (maximax). We demonstrate that, under these conditions, there exists a unique optimal combination of price and wage that maximizes the platform’s profit. This study reveals that the platform’s profit under the optimistic mindset is equivalent to that under certain demand or supply conditions. In response to criticisms that the maximin approach is excessively conservative, we show that the performance under this approach can reach up to 83% of the profit achieved under deterministic demand or supply, indicating that the maximin approach is less conservative than often claimed. To test the robustness of our main results, we extend the model to incorporate an additive structure for the demand or supply function, as well as the scenario where the demand or supply functions follow a uniform distribution. The analysis demonstrates that, under certain conditions, the platform’s profit under partial information can be up to 2.5 times higher than in the case of a uniform distribution.

Optimizing inventory management in the pharmaceutical industry relies significantly on accurate sales forecasting for chain drugstores. Sales predictions for these stores, however, are affected by data quality and temporal characteristics, which limit the effectiveness of traditional statistical, machine learning, and ensemble learning methods. To address these challenges, this study introduces a sales forecasting model called TS-LGBM, which utilizes a sliding window approach to preserve the sequential integrity of sales data and integrates neural networks with the Light Gradient Boosting Machine (LightGBM). By incorporating a self-attention mechanism into LightGBM, the TS-LGBM model aims to enhance predictive accuracy. The model’s efficacy is validated using the Rossman dataset from Kaggle, followed by a case study with actual data from Z-chain retail drugstores. This study further refines data by factoring in temporal characteristics of various drugs, the density of nearby drugstores within a specified radius, and regional attributes associated with each drugstore. To evaluate performance, five models—TS-LGBM, TS-XGB, LightGBM, XGBoost and LSTM—are compared experimentally. Findings indicate that TS-LGBM achieves superior prediction accuracy compared to the other models. This study is intended for practical applications, as accurate sales forecasts for chain drugstores can enhance supply chain management efficiency and reduce operational costs.