Journal of Big Data最新文献

In digital agriculture, accurate crop detection is fundamental to developing automated systems for efficient plantation management. For oil palm, the main challenge lies in developing robust models that perform well in different environmental conditions. This study addresses the feasibility of using GAN augmentation methods to improve palm detection models. For this purpose, drone images of young palms (< 5 year-old) from eight different estates were collected, annotated, and used to build a baseline detection model based on DETR. StyleGAN2 was trained on the extracted palms and then used to generate a series of synthetic palms, which were then inserted into tiles representing different environments. CycleGAN networks were trained for bidirectional translation between synthetic and real tiles, subsequently utilized to augment the authenticity of synthetic tiles. Both synthetic and real tiles were used to train the GAN-based detection model. The baseline model achieved precision and recall values of 95.8% and 97.2%. The GAN-based model achieved comparable result, with precision and recall values of 98.5% and 98.6%. In the challenge dataset 1 consisting older palms (> 5 year-old), both models also achieved similar accuracies, with baseline model achieving precision and recall of 93.1% and 99.4%, and GAN-based model achieving 95.7% and 99.4%. As for the challenge dataset 2 consisting of storm affected palms, the baseline model achieved precision of 100% but recall was only 13%. The GAN-based model achieved a significantly better result, with a precision and recall values of 98.7% and 95.3%. This result demonstrates that images generated by GANs have the potential to enhance the accuracies of palm detection models.

The question of whether artificial intelligence (AI) can surpass human capabilities is crucial in the application of AI in clinical medicine. To explore this, an interpretable deep learning (DL) model was developed to assess myopia status using retinal refraction maps obtained with a novel peripheral refractor. The DL model demonstrated promising performance, achieving an AUC of 0.9074 (95% CI 0.83–0.97), an accuracy of 0.8140 (95% CI 0.70–0.93), a sensitivity of 0.7500 (95% CI 0.51–0.90), and a specificity of 0.8519 (95% CI 0.68–0.94). Grad-CAM analysis provided interpretable visualization of the attention of DL model and revealed that the DL model utilized information from the central retina, similar to human readers. Additionally, the model considered information from vertical regions across the central retina, which human readers had overlooked. This finding suggests that AI can indeed surpass human capabilities, bolstering our confidence in the use of AI in clinical practice, especially in new scenarios where prior human knowledge is limited.

In recent years, mobile applications have proliferated across domains such as E-banking, Augmented Reality, E-Transportation, and E-Healthcare. These applications are often built using microservices, an architectural style where the application is composed of independently deployable services focusing on specific functionalities. Mobile devices cannot process these microservices locally, so traditionally, cloud-based frameworks using cost-efficient Virtual Machines (VMs) and edge servers have been used to offload these tasks. However, cloud frameworks suffer from extended boot times and high transmission overhead, while edge servers have limited computational resources. To overcome these challenges, this study introduces a Microservices Container-Based Mobile Edge Cloud Computing (MCBMEC) environment and proposes an innovative framework, Optimization Task Scheduling and Computational Offloading with Cost Awareness (OTSCOCA). This framework addresses Resource Matching, Task Sequencing, and Task Scheduling to enhance server utilization, reduce service latency, and improve service bootup times. Empirical results validate the efficacy of MCBMEC and OTSCOCA, demonstrating significant improvements in server efficiency, reduced service latency, faster service bootup times, and notable cost savings. These outcomes underscore the pivotal role of these methodologies in advancing mobile edge computing applications amidst the challenges of edge server limitations and traditional cloud-based approaches.

The success of newly established companies holds significant implications for community development and economic growth. However, startups often grapple with heightened vulnerability to market volatility, which can lead to early-stage failures. This study aims to predict startup success by addressing biases in existing predictive models. Previous research has examined external factors such as market dynamics and internal elements like founder characteristics.While such efforts have contributed to understanding success mechanisms, challenges persist, including predictor and learning data biases. This study proposes a novel approach by constructing independent variables using early-stage information, incorporating founder attributes, and mitigating class imbalance through generative adversarial networks (GAN). Our proposed model aims to enhance investment decision-making efficiency and effectiveness, offering a valuable decision support system for various venture capital funds.

Class imbalance is one of many problems of customer churn datasets. One of the common problems is class overlap, where the data have a similar instance between classes. The prediction task of customer churn becomes more challenging when there is class overlap in the data training. In this research, we suggested a hybrid method based on tabular GANs, called CTGAN-ENN, to address class overlap and imbalanced data in datasets of customers that churn. We used five different customer churn datasets from an open platform. CTGAN is a tabular GAN-based oversampling to address class imbalance but has a class overlap problem. We combined CTGAN with the ENN under-sampling technique to overcome the class overlap. CTGAN-ENN reduced the number of class overlaps by each feature in all datasets. We investigated how effective CTGAN-ENN is in each machine learning technique. Based on our experiments, CTGAN-ENN achieved satisfactory results in KNN, GBM, XGB and LGB machine learning performance for customer churn predictions. We compared CTGAN-ENN with common over-sampling and hybrid sampling methods, and CTGAN-ENN achieved outperform results compared with other sampling methods and algorithm-level methods with cost-sensitive learning in several machine learning algorithms. We provide a time consumption algorithm between CTGAN and CTGAN-ENN. CTGAN-ENN achieved less time consumption than CTGAN. Our research work provides a new framework to handle customer churn prediction problems with several types of imbalanced datasets and can be useful in real-world data from customer churn prediction.

Some of the most notable human behavioral palimpsests result from warfare and its durable traces in the form of defensive architecture and strategic infrastructure. For premodern periods, this architecture is often understudied at the large scale, resulting in a lack of appreciation for the enormity of the costs and impacts of military spending over the course of human history. In this article, we compare the information gleaned from the study of the fortified cities of the Early Historic period of the Indian subcontinent (c. 3rd century BCE to 4th century CE) with the precolonial medieval era (9-17th centuries CE). Utilizing in-depth archaeological and historical studies along with local sightings and citizen-science blogs to create a comprehensive data set and map series in a “big-data” approach that makes use of heterogeneous data sets and presence-absence criteria, we discuss how the architecture of warfare shifted from an emphasis on urban defense in the Early Historic period to an emphasis on territorial offense and defense in the medieval period. Many medieval fortifications are known from only local reports and have minimal identifying information but can still be studied in the aggregate using a least-shared denominator approach to quantification and mapping.

Subway button detection is paramount for passenger safety, yet the occurrence of inadvertent touches poses operational threats. Camera-based detection is indispensable for identifying touch occurrences, ascertaining person identity, and implementing scientific measures. Existing methods suffer from inaccuracies due to the small size of buttons, complex environments, and challenges such as occlusion. We present YOLOv8-DETR-P2-DCNv2-Dynamic-NWD-DA, which enhances occlusion awareness, reduces redundant annotations, and improves contextual feature extraction. The model integrates the RTDETRDecoder, P2 small target detection layer, DCNv2-Dynamic algorithm, and the NWD loss function for multiscale feature extraction. Dataset augmentation and the GAN algorithm refine the model, aligning feature distributions and enhancing precision by 6.5%, 5%, and 5.8% in precision, recall, and mAP50, respectively. These advancements denote significant improvements in key performance indicators.

This study investigates the causes and countermeasures of cybercrime vulnerabilities, specifically focusing on selected 16 Ethiopian university websites. This study uses a cybersecurity awareness survey, and automated vulnerability assessment and penetration testing (VAPT) technique tools, namely, Nmap, Nessus, and Vega, to identify potential security threats and vulnerabilities. The assessment was performed according to the ISO/IEC 27001 series of standards, ensuring a comprehensive and globally recognized approach to information security. The results of this study provide valuable insights into the current state of cybersecurity in Ethiopian universities and reveals a range of issues, from outdated software and poor password management to a lack of encryption and inadequate access control. Vega vulnerability assessment reports 11,286 total findings, and Nessus identified a total of 1749 vulnerabilities across all the websites of the institutions examined. Based on these findings, the study proposes counteractive measures tailored to the specific needs of each identified defect. These recommendations aim to strengthen the security posture of the university websites, thereby protecting sensitive data and maintaining the trust of students, staff, and other stakeholders. The study emphasizes the need for proactive cybersecurity measures in the realm of higher education and presents a strategic plan for universities to improve their digital security.

Crude oil is an essential energy source that affects international trade, transportation, and manufacturing, highlighting its importance to the economy. Its future price prediction affects consumer prices and the energy markets, and it shapes the development of sustainable energy. It is essential for financial planning, economic stability, and investment decisions. However, reaching a reliable future prediction is an open issue because of its high volatility. Furthermore, many state-of-the-art methods utilize signal decomposition techniques, which can lead to increased prediction time. In this paper, a model called K-means-dense-sparse-dense long short-term memory (K-means-DSD-LSTM) is proposed, which has three main training phrases for crude oil price forecasting. In the first phase, the DSD-LSTM model is trained. Afterwards, the training part of the data is clustered using the K-means algorithm. Finally, a copy of the trained DSD-LSTM model is fine-tuned for each obtained cluster. It helps the models predict that cluster better while they are generalizing the whole dataset quite well, which diminishes overfitting. The proposed model is evaluated on two famous crude oil benchmarks: West Texas Intermediate (WTI) and Brent. Empirical evaluations demonstrated the superiority of the DSD-LSTM model over the K-means-LSTM model. Furthermore, the K-means-DSD-LSTM model exhibited even stronger performance. Notably, the proposed method yielded promising results across diverse datasets, achieving competitive performance in comparison to existing methods, even without employing signal decomposition techniques.

U-net, a fully convolutional network-based image segmentation method, has demonstrated widespread adaptability in the crack segmentation task. The combination of the semantically dissimilar features of the encoder (shallow layers) and the decoder (deep layers) in the skip connections leads to blurry features map and leads to undesirable over- or under-segmentation of target regions. Additionally, the shallow architecture of the U-Net model prevents the extraction of more discriminatory information from input images. This paper proposes a Residual Sharp U-Net (RS-Net) architecture for crack segmentation and severity assessment in pavement surfaces to address these limitations. The proposed architecture uses residual block in the U-Net model to extract a more insightful representation of features. In addition to that, a sharpening kernel filter is used instead of plain skip connections to generate a fine-tuned encoder features map before combining it with decoder features maps to reduce the dissimilarity between them and smoothes artifacts in the network layers during early training. The proposed architecture is also integrated with various morphological operations to assess the severity of cracks and categorize them into hairline, medium, and severe labels. Experiments results demonstrated that the RS-Net model has promising segmentation performance, outperforming earlier U-Net variations on testing data for crack segmentation and severity assessment, with a promising accuracy (>0.97)