Imputation of missing sensor-collected data is often an important step prior to machine learning and statistical data analysis. One particular data imputation challenge is filling large data gaps when the only related data comes from the same sensor station. In this paper, we propose a framework to improve the popular multivariate imputation by chained equations (MICE) method for dealing with missing data. One key strategy we use to improve model accuracy is to reshape the original sensor data to leverage the correlation between the missing data and the observed data. We demonstrate our framework using data from continuous water quality monitoring stations in Vermont. Because of possible irregularly spaced peaks throughout the time series, the reshaped data is split into extreme and normal values and two MICE models are built. We also recommend that sensor-collected data should be transformed to meet the machine learning model assumptions. According to our experimental results, these strategies can improve MICE data imputation model accuracy at least 23% for large data gaps based on $\text {R}^{{2}}$ values and are promising to be applied for other data imputation algorithms.

Low-cost monitors for measuring airborne contaminants have gained popularity due to their affordability, portability, and ease of use. However, they often exhibit significant biases compared to high-cost reference instruments. For optimal accuracy, these monitors require calibration and validation in their specific environment using expensive reference instruments, which are often scarce and costly. This study proposes machine-learning calibration methods that utilize a single high-cost instrument as an active reference to improve the accuracy of large networks of low-cost monitors. Three machine learning models-linear regression, random forest, and Gradient Boosting Regression (GBR)-were employed. The proposed approach was tested in a controlled chamber under two conditions: environmental simulations with salt- and dust-based aerosols and occupational settings using three electronic cigarette (ECIG) brands. The study involved thirty low-cost GeoAir2 monitors, divided into ten groups of three. Initially, all groups were collocated with a high-cost monitor using Aerosol A to develop prediction and regression models. These models, along with intrinsic error measurements from one group, were then applied to improve data accuracy for the remaining groups using Aerosol B. The results demonstrated substantial improvements in accuracy, with r² values ranging from 0.91 to 1.00 and RMSE reductions of up to 88 %, depending on the model and aerosol type. GBR consistently provided the highest accuracy and performance, particularly for complex, nonlinear patterns, while linear regression offered a faster, computationally efficient alternative suitable for less demanding scenarios. Random forest models performed moderately well, balancing accuracy and complexity. These methods provide a scalable and cost-effective solution for deploying networked low-cost sensors. Further research is needed to validate these findings in outdoor environments with meteorological and spatial influences, and indoor occupational settings where humidity effects may play a role.

Image matching plays a critical role in a wide range of computer vision applications, including object recognition, 3D reconstruction, aiming-point and six-degree-of-freedom detection for aiming devices, and video surveillance. Over the past three decades, image-matching algorithms and techniques have evolved significantly, from handcrafted feature extraction algorithms to modern approaches powered by deep learning neural networks and attention mechanisms. This paper provides a comprehensive review of image-matching techniques, aiming to offer researchers valuable insights into the evolving landscape of this field. It traces the historical development of feature-based methods and examines the transition to neural network-based approaches that leverage large-scale data and learned representations. Additionally, this paper discusses the current state of the field, highlighting key algorithms, benchmarks, and real-world applications. Furthermore, this study introduces some recent contributions to this area and outlines promising directions for future research, including H-matrix optimization, LoFTR model speedup, and performance improvements. It also identifies persistent challenges such as robustness to viewpoint and illumination changes, scalability, and matching under extreme conditions. Finally, this paper summarizes future trends for research and development in this field.

In this paper, we investigate the existence and uniqueness of solutions for nonlinear quadratic iterative equations in the sense of the Caputo fractional derivative with different boundary conditions. Under a one-sided-Lipschitz condition on the nonlinear term, the existence and uniqueness of a solution for the boundary value problems of Caputo fractional iterative equations with arbitrary order is demonstrated by applying the Leray-Schauder fixed point theorem and topological degree theory, where the solution for the case of fractional order greater than 1 is monotonic. Then, the existence and uniqueness of a solution for the period and integral boundary value problems of Caputo fractional quadratic iterative equations in $ R^N $ are also demonstrated. Furthermore, the well posedness of the control problem of a nonlinear iteration system with a disturbance is established by applying set-valued theory, and the existence of solutions for a neural network iterative system is guaranteed. As an application, an example is provided at the end.

Adults who work in the Central Appalachian region of the United States (U.S.) are disproportionately affected by Chronic Obstructive Pulmonary Disease (COPD). While there is a socio-demographic profile of adults with COPD who are at increased risk for physical and mental distress, the risk factors that uniquely affect the health-related quality of life (HRQoL) of Central Appalachian workers with COPD are unknown. Therefore, we conducted a latent class analysis of 2016 and 2017 Behavioral Risk Factor Surveillance System data from 1326 currently employed adults with COPD living in four U.S. states (KY, NC, TN, and WV) within the Central Appalachian Region. Drawing from the social ecological model, we identified associations between theoretically informed risk indicators—comorbid health conditions, substance use and abuse, and limited access to healthcare—on three HRQoL variables, including infrequent (0–13 days) or frequent (≥14 days) physical distress, mental distress, and limited activity due to poor health over the past 30 days. Workers at high risk for comorbid conditions reported more frequent physical distress, mental distress, and activity limitations as compared to those at low risk. Workers reporting difficulty accessing healthcare were no more likely to report physical or mental distress when compared to workers with adequate access to healthcare; however, those with limited healthcare access did report more frequent activity limitation due to poor health. Interestingly, workers with COPD at high risk for substance use and abuse were no more likely to report poor HRQoL outcomes compared to those at low risk. Our findings have important implications for addressing indicators of poor health among Central Appalachian workers with COPD, especially those living with multiple comorbidities.