Explore the vast range of titles available.

Tiny defect detection (TDD) which aims to perform the quality control of printed circuit boards (PCBs) is a basic and essential task in the production of most electronic products. Though significant progress has been made in PCB defect detection, traditional methods are still difficult to cope with the complex and diverse PCBs. To deal with these problems, this article proposes a tiny defect detection network (TDD-Net) to improve performance for PCB defect detection. In this method, the inherent multi-scale and pyramidal hierarchies of deep convolutional networks are exploited to construct feature pyramids. Compared with existing approaches, the TDD-Net has three novel changes. First, reasonable anchors are designed by using k-means clustering. Second, TDD-Net strengthens the relationship of feature maps from different levels and benefits from low-level structural information, which is suitable for tiny defect detection. Finally, considering the small and imbalance dataset, online hard example mining is adopted in the whole training phase in order to improve the quality of region-of-interest (ROI) proposals and make more effective use of data information. Quantitative results on the PCB defect dataset show that the proposed method has better portability and can achieve 98.90% mAP, which outperforms the state-of-arts. The code will be publicly available.

Big data analysis requires the presence of large computing powers, which is not always feasible. And so, it became necessary to develop new clustering algorithms capable of such data processing. This study proposes a new parallel clustering algorithm based on the k-means algorithm. It significantly reduces the exponential growth of computations. The proposed algorithm splits a dataset into batches while preserving the characteristics of the initial dataset and increasing the clustering speed. The idea is to define cluster centroids, which are also clustered, for each batch. According to the obtained centroids, the data points belong to the cluster with the nearest centroid. Real large datasets are used to conduct the experiments to evaluate the effectiveness of the proposed approach. The proposed approach is compared with k-means and its modification. The experiments show that the proposed algorithm is a promising tool for clustering large datasets in comparison with the k-means algorithm.

Vector Quantization (VQ) is a clustering problem in the fields of signal processing, source coding, information theory etc. Taking advantage of recent advances in the field of deep neural networks, this paper investigates the performance between VQ clustering problems and deep neural networks. A k‐means‐based deep network architecture for VQ is presented to solve clustering problems. By applying the deep learning implementation of convergence optimization, a clustering neural network (algorithm) for the purpose of VQ is proposed. In practice, the proposed network quantifies the vectors over a set of stacked neural layers, overcoming the exponential complexity problem of VQ methods by trainable parameters. Experiments show that the work can improve the results without human intervention, and outperforms traditional clustering methods modified for VQ. A learnable clustering neural network is proposed for VQ tasks. The clustering center is defined as the training parameters that can achieve the traditional algorithm by deep learning approach.

This paper discusses the possibility of recognizing and predicting user activities in the IoT (Internet of Things) based smart environment. The activity recognition is usually done through two steps: activity pattern clustering and activity type decision. Although many related works have been suggested, they had some limited performance because they focused only on one part between the two steps. This paper tries to find the best combination of a pattern clustering method and an activity decision algorithm among various existing works. For the first step, in order to classify so varied and complex user activities, we use a relevant and efficient unsupervised learning method called the K-pattern clustering algorithm. In the second step, the training of smart environment for recognizing and predicting user activities inside his/her personal space is done by utilizing the artificial neural network based on the Allen’s temporal relations. The experimental results show that our combined method provides the higher recognition accuracy for various activities, as compared with other data mining classification algorithms. Furthermore, it is more appropriate for a dynamic environment like an IoT based smart home.

In southeastern North America, Indigenous potters and woodworkers carved complex, primarily abstract, designs into wooden pottery paddles, which were subsequently used to thin the walls of hand-built, clay vessels. Original paddle designs carry rich historical and cultural information, but pottery paddles from ancient times have not survived. Archaeologists have studied design fragments stamped on sherds to reconstruct complete or nearly complete designs, which is extremely laborious and time-consuming. In Snowvision, we aim to develop computer vision methods to assist archaeologists to accomplish this goal more efficiently and effectively. For this purpose, we identify and study three computer vision tasks: (1) extracting curve structures stamped on pottery sherds; (2) matching sherds to known designs; (3) clustering sherds with unknown designs. Due to the noisy, highly fragmented, composite-curve patterns, each task poses unique challenges to existing methods. To solve them, we propose (1) a weakly-supervised CNN-based curve structure segmentation method that takes only curve skeleton labels to predict full curve masks; (2) a patch-based curve pattern matching method to address the problem of partial matching in terms of noisy binary images; (3) a curve pattern clustering method consisting of pairwise curve matching, graph partitioning and sherd stitching. We evaluate the proposed methods on a set of collected sherds and extensive experimental results show the effectiveness of the proposed algorithms.

Clustering algorithm is the primary technology used in target clustering and group status analysis which are key features of the Unmanned Aerial Vehicles (UAVs) control system. Due to variable application environment, the stability of the algorithm in the UAVs control system needs to be considered. K‐means clustering is a widely used method in intelligent systems. However, K‐means algorithm is susceptible to the local optimum due to the influence of the initial centroid. For this problem, the predecessors have proposed various effective solutions. These algorithms perform better on real and large‐scale datasets, but they are unable to achieve optimum results with unbalanced datasets. Herein, a simpler and more effective algorithm for seed initialization is proposed, it has a better accuracy rate than the alternative algorithms.Moreover, after running tests multiple times with each algorithm independently, it has the highest stability and the lowest overall volatility. With unbalanced datasets, the proposed algorithm performs significantly better than several other algorithms and therefore can solve the problems that other algorithms have with unbalanced datasets. K‐means clustering is a widely used method in intelligent systems. However, K‐means algorithm is susceptible to the local optimum due to the influence of the initial centroid. Herein, a simpler and more effective algorithm for seed initialization is proposed. With unbalanced datasets, the proposed algorithm performs significantly better than several other algorithms and therefore can solve the problems that other algorithms have with unbalanced datasets.

Automating the quality assessment of Software Requirement Specification poses major challenges related to the need for advanced algorithms to extract the SRS quality features, interpret the context of the features, formulate accurate assessment metrics, and document the shortcomings as well as possible improvements. In the existing methods, such as Reconstructed Automated Requirement Measurement, and Rendex, some major processes are still handled offline by humans (semi‐automated) or encompass automating the measurement of a few quality attributes due to the mentioned challenges. This paper addressed this gap and proposed an Automated Quality Assessment of SRS (AQA‐SRS) framework to assess the SRS documents by automatically extracting features related to 11 quality attributes through a deep analysis of the SRS textual content. Also, it constructs a flexible platform that is able to minimize the human expert’s role in the SRS assessment. The AQA‐SRS framework integrates Natural Language Processing, K‐means, Multi‐agent, and Case‐Based Reasoning. The AQA‐SRS framework is evaluated by processing two standard SRS datasets and comparing the results with state‐of‐the‐art methods and analysis by software engineering experts. The results show that the AQA‐SRS framework effectively assesses the tested SRS documents and achieves a 78% total agreement with the tested methods and software engineering experts. This paper proposed an Automated Quality Assessment of SRS (AQA‐SRS) framework by integrating four popular methods which are; NLP, K‐means, MAS, and CBR to assess the quality of SRS documents. The NLP utilize for feature extraction, K‐means for features clustering, MAS for interactive assessment and feature selection decision, and CBR for managing the entire assessment process.

Classical radial basis function network (RBFN) is widely used to process the non-linear separable data sets with the introduction of activation functions. However, the setting of parameters for activation functions is random and the distribution of patterns is not taken into account. To process this issue, some scholars introduce the kernel clustering into the RBFN so that the clustering results are related to the parameters about activation functions. On the base of the original kernel clustering, this study further discusses the influence of kernel clustering on an RBFN when the setting of kernel clustering is changing. The changing involves different kernel-clustering ways [bubble sort (BS) and escape nearest outlier (ENO)], multiple kernel-clustering criteria (static and dynamic) etc. Experimental results validate that with the consideration of distribution of patterns and the changes of setting of kernel clustering, the performance of an RBFN is improved and is more feasible for corresponding data sets. Moreover, though BS always costs more time than ENO, it still brings more feasible clustering results. Furthermore, dynamic criterion always cost much more time than static one, but kernel number derived from dynamic criterion is fewer than the one from static.

Crime against women (CAW) persists as an enduring and disconcerting societal issue, encompassing a range of physical, sexual, and emotional violence. This study examines the CAW in West Bengal, India, by drawing on recent official data from the National Crime Records Bureau. The data covers the period from 2016 to 2021, providing a robust foundation for assessing these crimes' spatial and temporal patterns. Six distinct categories of CAW have been selected for comprehensive analysis: dowry death, cruelty by husband or relatives, kidnapping and abduction, rape, assault on women, and insult. Employing Geographical Information Systems and spatial analysis techniques, this research scrutinizes the geographical distribution of these crimes across diverse districts within West Bengal. Through meticulous examination of data from 2016 to 2021, the study aims to identify spatial and temporal variations exhibited by these crimes over time. Spatial autocorrelation analysis utilizing Moran's I technique is employed to unveil clustering patterns associated with specific types of CAW. Furthermore, the Getis-Ord Gi* technique is utilized to discern hotspot and coldspot patterns, illuminating those districts characterized by concentrated occurrences of similar crimes targeting women. The findings underscore that certain district, such as North 24 Parganas, South 24 Parganas, Nadia, and Murshidabad, witness a higher incidence of overall CAW than other districts. Understanding these crimes' spatial and temporal dynamics is crucial for effective policymaking, prevention strategies, and resource allocation. The study emphasizes the importance of evidence-based approaches in addressing CAW and creating safer environments for women in West Bengal.

Electrical load time series exhibit strong heterogeneity across daily patterns driven by calendar effects and behavioral variability, leading many forecasting models to favor dominant weekday profiles while degrading on weekends, holidays, and transition days. This paper proposes an imbalance-aware spatiotemporal forecasting framework via a cluster-conditioned state space model. Daily load patterns are identified via time-series clustering and incorporated as conditioning covariates within a sequence-continuous selective state space models (Mamba), preserving temporal coherence without explicit sequence partitioning. A cluster-weighted training objective further mitigates pattern imbalance while avoiding future-information leakage. The resulting cluster-conditioned Time Series Mamba (TSMamba) consistently improves forecasting robustness across both frequent and infrequent profiles, achieving weighted absolute percentage error (WAPE) reductions of approximately 15% on weekdays, 42% on weekends, and 39% on holidays relative to the vanilla TSMamba, with similar gains in mean absolute error (MAE) and coefficient of variation of the root mean square error (CVRMSE). These results demonstrate that conditioning state dynamics on latent load patterns yields stable and computationally efficient short-term load forecasts under profile transitions.