Frontiers in Big Data最新文献

Introduction: Recently, Google introduced Pathways as its next-generation AI architecture. Pathways must address three critical challenges: learning one general model for several continuous tasks, ensuring tasks can leverage each other without forgetting old tasks, and learning from multi-modal data such as images and audio. Additionally, Pathways must maintain sparsity in both learning and deployment. Current lifelong multi-task learning approaches are inadequate in addressing these challenges.

Methods: To address these challenges, we propose SEN, a Sparse and Expandable Network. SEN is designed to handle multiple tasks concurrently by maintaining sparsity and enabling expansion when new tasks are introduced. The network leverages multi-modal data, integrating information from different sources while preventing interference between tasks.

Results: The proposed SEN model demonstrates significant improvements in multi-task learning, successfully managing task interference and forgetting. It effectively integrates data from various modalities and maintains efficiency through sparsity during both the learning and deployment phases.

Discussion: SEN offers a straightforward yet effective solution to the limitations of current lifelong multi-task learning methods. By addressing the challenges identified in the Pathways architecture, SEN provides a promising approach for developing AI systems capable of learning and adapting over time without sacrificing performance or efficiency.

Time series data are recorded in various sectors, resulting in a large amount of data. However, the continuity of these data is often interrupted, resulting in periods of missing data. Several algorithms are used to impute the missing data, and the performance of these methods is widely varied. Apart from the choice of algorithm, the effective imputation depends on the nature of missing and available data. We conducted extensive studies using different types of time series data, specifically heart rate data and power consumption data. We generated the missing data for different time spans and imputed using different algorithms with binned data of different sizes. The performance was evaluated using the root mean square error (RMSE) metric. We observed a reduction in RMSE when using binned data compared to the entire dataset, particularly in the case of the expectation-maximization (EM) algorithm. We found that RMSE was reduced when using binned data for 1-, 5-, and 15-min missing data, with greater reduction observed for 15-min missing data. We also observed the effect of data fluctuation. We conclude that the usefulness of binned data depends precisely on the span of missing data, sampling frequency of the data, and fluctuation within data. Depending on the inherent characteristics, quality, and quantity of the missing and available data, binned data can impute a wide variety of data, including biological heart rate data derived from the Internet of Things (IoT) device smartwatch and non-biological data such as household power consumption data.

Differential privacy (DP) has been in the public spotlight since the announcement of its use in the 2020 U.S. Census. While DP algorithms have substantially improved the confidentiality protections provided to Census respondents, concerns have been raised about the accuracy of the DP-protected Census data. The extent to which the use of DP distorts the ability to draw inferences that drive policy about small-populations, especially marginalized communities, has been of particular concern to researchers and policy makers. After all, inaccurate information about marginalized populations can often engender policies that exacerbate rather than ameliorate social inequities. Consequently, computer science experts have focused on developing mechanisms that help achieve equitable privacy, i.e., mechanisms that mitigate the data distortions introduced by privacy protections to ensure equitable outcomes and benefits for all groups, particularly marginalized groups. Our paper extends the conversation on equitable privacy by highlighting the importance of inclusive communication in ensuring equitable outcomes for all social groups through all the stages of deploying a differentially private system. We conceptualize Equitable DP as the design, communication, and implementation of DP algorithms that ensure equitable outcomes. Thus, in addition to adopting computer scientists' recommendations of incorporating equity parameters within DP algorithms, we suggest that it is critical for an organization to also facilitate inclusive communication throughout the design, development, and implementation stages of a DP algorithm to ensure it has an equitable impact on social groups and does not hinder the redressal of social inequities. To demonstrate the importance of communication for Equitable DP, we undertake a case study of the process through which DP was adopted as the newest disclosure avoidance system for the 2020 U.S. Census. Drawing on the Inclusive Science Communication (ISC) framework, we examine the extent to which the Census Bureau's communication strategies encouraged engagement across the diverse groups of users that employ the decennial Census data for research and policy making. Our analysis provides lessons that can be used by other government organizations interested in incorporating the Equitable DP approach in their data collection practices.

Introduction: "Data scientists" quickly became ubiquitous, often infamously so, but they have struggled with the ambiguity of their novel role. This article studies data science's collective definition on Twitter.

Methods: The analysis responds to the challenges of studying an emergent case with unclear boundaries and substance through a cultural perspective and complementary datasets ranging from 1,025 to 752,815 tweets. It brings together relations between accounts that tweeted about data science, the hashtags they used, indicating purposes, and the topics they discussed.

Results: The first results reproduce familiar commercial and technical motives. Additional results reveal concerns with new practical and ethical standards as a distinctive motive for constructing data science.

Discussion: The article provides a sensibility for local meaning in usually abstract datasets and a heuristic for navigating increasingly abundant datasets toward surprising insights. For data scientists, it offers a guide for positioning themselves vis-à-vis others to navigate their professional future.

The advent of the digital era has transformed E-commerce platforms into critical tools for industry, yet traditional recommendation systems often fall short in the specialized context of the electric power industry. These systems typically struggle with the industry's unique challenges, such as infrequent and high-stakes transactions, prolonged decision-making processes, and sparse data. This research has developed a novel recommendation engine tailored to these specific conditions, such as to handle the low frequency and long cycle nature of Business-to-Business (B2B) transactions. This approach includes algorithmic enhancements to better process and interpret the limited data available, and data pre-processing techniques designed to enrich the sparse datasets characteristic of this industry. This research also introduces a methodological innovation that integrates multi-dimensional data, combining user E-commerce activities, product specifics, and essential non-tendering information. The proposed engine employs advanced machine learning techniques to provide more accurate and relevant recommendations. The results demonstrate a marked improvement over traditional models, offering a more robust and effective tool for facilitating B2B transactions in the electric power industry. This research not only addresses the sector's unique challenges but also provides a blueprint for adapting recommendation systems to other industries with similar B2B characteristics.

Low-rank tensor completion (LRTC), which aims to complete missing entries from tensors with partially observed terms by utilizing the low-rank structure of tensors, has been widely used in various real-world issues. The core tensor nuclear norm minimization (CTNM) method based on Tucker decomposition is one of common LRTC methods. However, the CTNM methods based on Tucker decomposition often have a large computing cost due to the fact that the general factor matrix solving technique involves multiple singular value decompositions (SVDs) in each loop. To address this problem, this article enhances the method and proposes an effective CTNM method based on thin QR decomposition (CTNM-QR) with lower computing complexity. The proposed method extends the CTNM by introducing tensor versions of the auxiliary variables instead of matrices, while using the thin QR decomposition to solve the factor matrix rather than the SVD, which can save the computational complexity and improve the tensor completion accuracy. In addition, the CTNM-QR method's convergence and complexity are analyzed further. Numerous experiments in synthetic data, real color images, and brain MRI data at different missing rates demonstrate that the proposed method not only outperforms in terms of completion accuracy and visualization, but also conducts more efficiently than most state-of-the-art LRTC methods.

The k-nearest neighbors (KNN) regression method, known for its nonparametric nature, is highly valued for its simplicity and its effectiveness in handling complex structured data, particularly in big data contexts. However, this method is susceptible to overfitting and fit discontinuity, which present significant challenges. This paper introduces the random kernel k-nearest neighbors (RK-KNN) regression as a novel approach that is well-suited for big data applications. It integrates kernel smoothing with bootstrap sampling to enhance prediction accuracy and the robustness of the model. This method aggregates multiple predictions using random sampling from the training dataset and selects subsets of input variables for kernel KNN (K-KNN). A comprehensive evaluation of RK-KNN on 15 diverse datasets, employing various kernel functions including Gaussian and Epanechnikov, demonstrates its superior performance. When compared to standard KNN and the random KNN (R-KNN) models, it significantly reduces the root mean square error (RMSE) and mean absolute error, as well as improving R-squared values. The RK-KNN variant that employs a specific kernel function yielding the lowest RMSE will be benchmarked against state-of-the-art methods, including support vector regression, artificial neural networks, and random forests.

With the increasing popularity of Graph Neural Networks (GNNs) for predictive tasks on graph structured data, research on their explainability is becoming more critical and achieving significant progress. Although many methods are proposed to explain the predictions of GNNs, their focus is mainly on "how to generate explanations." However, other important research questions like "whether the GNN explanations are inaccurate," "what if the explanations are inaccurate," and "how to adjust the model to generate more accurate explanations" have gained little attention. Our previous GNN Explanation Supervision (GNES) framework demonstrated effectiveness on improving the reasonability of the local explanation while still keep or even improve the backbone GNNs model performance. In many applications instead of per sample explanations, we need to find global explanations which are reasonable and faithful to the domain data. Simply learning to explain GNNs locally is not an optimal solution to a global understanding of the model. To improve the explainability power of the GNES framework, we propose the Global GNN Explanation Supervision (GGNES) technique which uses a basic trained GNN and a global extension of the loss function used in the GNES framework. This GNN creates local explanations which are fed to a Global Logic-based GNN Explainer, an existing technique that can learn the global Explanation in terms of a logic formula. These two frameworks are then trained iteratively to generate reasonable global explanations. Extensive experiments demonstrate the effectiveness of the proposed model on improving the global explanations while keeping the performance similar or even increase the model prediction power.

Tuberculosis (TB) is a chronic and pathogenic disease that leads to life-threatening situations like death. Many people have been affected by TB owing to inaccuracy, late diagnosis, and deficiency of treatment. The early detection of TB is important to protect people from the severity of the disease and its threatening consequences. Traditionally, different manual methods have been used for TB prediction, such as chest X-rays and CT scans. Nevertheless, these approaches are identified as time-consuming and ineffective for achieving optimal results. To resolve this problem, several researchers have focused on TB prediction. Conversely, it results in a lack of accuracy, overfitting of data, and speed. For improving TB prediction, the proposed research employs the Selection Focal Fusion (SFF) block in the You Look Only Once v8 (YOLOv8, Ultralytics software company, Los Angeles, United States) object detection model with attention mechanism through the Kaggle TBX-11k dataset. The YOLOv8 is used for its ability to detect multiple objects in a single pass. However, it struggles with small objects and finds it impossible to perform fine-grained classifications. To evade this problem, the proposed research incorporates the SFF technique to improve detection performance and decrease small object missed detection rates. Correspondingly, the efficacy of the projected mechanism is calculated utilizing various performance metrics such as recall, precision, F1Score, and mean Average Precision (mAP) to estimate the performance of the proposed framework. Furthermore, the comparison of existing models reveals the efficiency of the proposed research. The present research is envisioned to contribute to the medical world and assist radiologists in identifying tuberculosis using the YOLOv8 model to obtain an optimal outcome.

Introduction: In response to the increasing prevalence of electronic medical records (EMRs) stored in databases, healthcare staff are encountering difficulties retrieving these records due to their limited technical expertise in database operations. As these records are crucial for delivering appropriate medical care, there is a need for an accessible method for healthcare staff to access EMRs.

Methods: To address this, natural language processing (NLP) for Text-to-SQL has emerged as a solution, enabling non-technical users to generate SQL queries using natural language text. This research assesses existing work on Text-to-SQL conversion and proposes the MedT5SQL model specifically designed for EMR retrieval. The proposed model utilizes the Text-to-Text Transfer Transformer (T5) model, a Large Language Model (LLM) commonly used in various text-based NLP tasks. The model is fine-tuned on the MIMICSQL dataset, the first Text-to-SQL dataset for the healthcare domain. Performance evaluation involves benchmarking the MedT5SQL model on two optimizers, varying numbers of training epochs, and using two datasets, MIMICSQL and WikiSQL.

Results: For MIMICSQL dataset, the model demonstrates considerable effectiveness in generating question-SQL pairs achieving accuracy of 80.63%, 98.937%, and 90% for exact match accuracy matrix, approximate string-matching, and manual evaluation, respectively. When testing the performance of the model on WikiSQL dataset, the model demonstrates efficiency in generating SQL queries, with an accuracy of 44.2% on WikiSQL and 94.26% for approximate string-matching.

Discussion: Results indicate improved performance with increased training epochs. This work highlights the potential of fine-tuned T5 model to convert medical-related questions written in natural language to Structured Query Language (SQL) in healthcare domain, providing a foundation for future research in this area.