High-level Analytics Research Articles

Missing Data in Sport Science: A Didactic Example Using Wearables in American Football

Data are often recorded from athletes to make decisions regarding the mitigation of injuries or the enhancement of performance. However, data collection in the real-world is difficult, and it is common for data to be missing from a particular training session due to equipment malfunction, athlete non-compliance, etc. The statistical community has long recognized that proper handling of missing data is vital to unbiased analyses and decision making, yet most dashboards in sport science and medicine do not recognize the issues introduced by missing data and practitioners are largely unaware that their displays are conveying biased information. The goal of this leading article is to show how real-world data can violate the 'missing completely at random' assumption in an American Football example and then demonstrate some potential imputation solutions which appear to maintain the underlying properties of the data in the presence of missingness. Whether data are aggregated on a dashboard as simple histograms and averages or with higher-level analytics, violating the 'missing completely at random' assumption results in a biased dashboard. Practitioners need to insist that dashboard developers perform missing data analyses and impute data as needed so valid data-driven decisions can be made.

Abstracting Data in Distributed Ledger Systems for Higher Level Analytics and Visualizations

By design, distributed ledger technologies persist low-level data, which makes conducting complex business analysis of the recorded operations challenging. Existing blockchain visualization and analytics tools such as block explorers tend to rely on this low-level data and complex interfacing to provide an enriched level of analytics. The ability to derive richer analytics could be improved through the availability of a higher level abstraction of the data. This article proposes an abstraction layer architecture that enables the design of high-level analytics of distributed ledger systems and the decentralized applications that run on top. Based on the analysis of existing initiatives and identification of the relevant user requirements, this work aims to establish key insights and specifications to improve the auditability and intuitiveness of distributed ledger systems by leveraging the development of future user interfaces. To illustrate the benefits offered by the proposed abstraction layer architecture, a regulated sector use case is explored.

JEDE: Universal Jersey Number Detector for Sports

The rapid progress in deep learning-based computer vision has opened unprecedented possibilities in computing various high-level analytics for sports. Artificial intelligence techniques such as predictive analysis, automatic highlight generation, and assistant coaching have been applied to improve performance and decision-making for teams and players. To perform any high-level analysis from a game match, collecting the locations (where) and identities (who) of players is crucial and challenging. In this paper, a universal JErsey number DEtector (JEDE) for player identification is presented that predicts players’ bounding boxes and keypoints, along with bounding boxes and classes of jersey digits and numbers in an end-to-end manner. Instead of generating digit proposals from pre-defined anchors, JEDE predicts more robust proposals guided by players’ features and pose estimation. Moreover, a dataset is collected from soccer and basketball matches with annotations on players’ bounding boxes and body keypoints, and jersey digits’ bounding boxes and labels. Extensive experimental results and ablation studies on the collected dataset show that the proposed method outperforms the state-of-the-art methods by a large margin. Both quantitative and qualitative results also demonstrate JEDE’s superior practicality and generalizability over different sports.

A universal sample-based background subtraction method for traffic surveillance videos

Background subtraction in traffic surveillance videos plays a crucial role in many high-level analytics and applications. Although great progress has been made in background subtraction, there are still many challenges in real traffic surveillance circumstances, such as camouflaged vehicles and dynamic background. This paper proposes a universal and accurate sampled-based vehicle detection method. The proposed method uses the color and Haar features to construct the background model, which can increase the accuracy of vehicle detection for camouflaged vehicles and is robust to low visibility. Besides, to reduce the incorrect samples of the initialized background model from a single image, the samples of the initial background model are randomly chosen from a similar candidate set. Furthermore, a novel random strategy is proposed to update the background pixel itself, while a combination update strategy with certainty and randomness is adopted for its neighborhood. This updating mechanism speeds up the suppression of ghosts in the background model. Experimental results verify the excellent behavior of our proposed method when compared to other mainstream methods.

Loginson: a transform and load system for very large-scale log analysis in large IT infrastructures

Nowadays, most systems and applications produce log records that are useful for security and monitoring purposes such as debugging programming errors, checking system status, and detecting configuration problems or even attacks. To this end, a log repository becomes necessary whereby logs can be accessed and visualized in a timely manner. This paper presents Loginson, a high-performance log centralization system for large-scale log collection and processing in large IT infrastructures. Besides log collection, Loginson provides high-level analytics through a visual interface for the purpose of troubleshooting critical incidents. We note that Loginson outperforms all of the other log centralization solutions by taking full advantage of the vertical scalability, and therefore decreasing Capital Expenditure (CAPEX) and Operating Expense (OPEX) costs for deployment scenarios with a huge volume of log data.

SuBSENSE: a universal change detection method with local adaptive sensitivity.

Foreground/background segmentation via change detection in video sequences is often used as a stepping stone in high-level analytics and applications. Despite the wide variety of methods that have been proposed for this problem, none has been able to fully address the complex nature of dynamic scenes in real surveillance tasks. In this paper, we present a universal pixel-level segmentation method that relies on spatiotemporal binary features as well as color information to detect changes. This allows camouflaged foreground objects to be detected more easily while most illumination variations are ignored. Besides, instead of using manually set, frame-wide constants to dictate model sensitivity and adaptation speed, we use pixel-level feedback loops to dynamically adjust our method's internal parameters without user intervention. These adjustments are based on the continuous monitoring of model fidelity and local segmentation noise levels. This new approach enables us to outperform all 32 previously tested state-of-the-art methods on the 2012 and 2014 versions of the ChangeDetection.net dataset in terms of overall F-Measure. The use of local binary image descriptors for pixel-level modeling also facilitates high-speed parallel implementations: our own version, which used no low-level or architecture-specific instruction, reached real-time processing speed on a midlevel desktop CPU. A complete C++ implementation based on OpenCV is available online.

Approximate Algorithms for Computing Spatial Distance Histograms with Accuracy Guarantees

Particle simulation has become an important research tool in many scientific and engineering fields. Data generated by such simulations impose great challenges to database storage and query processing. One of the queries against particle simulation data, the spatial distance histogram (SDH) query, is the building block of many high-level analytics, and requires quadratic time to compute using a straightforward algorithm. Previous work has developed efficient algorithms that compute exact SDHs. While beating the naive solution, such algorithms are still not practical in processing SDH queries against large-scale simulation data. In this paper, we take a different path to tackle this problem by focusing on approximate algorithms with provable error bounds. We first present a solution derived from the aforementioned exact SDH algorithm, and this solution has running time that is unrelated to the system size N. We also develop a mathematical model to analyze the mechanism that leads to errors in the basic approximate algorithm. Our model provides insights on how the algorithm can be improved to achieve higher accuracy and efficiency. Such insights give rise to a new approximate algorithm with improved time/accuracy tradeoff. Experimental results confirm our analysis.

Informatics in Radiology: Automated Web-based Graphical Dashboard for Radiology Operational Business Intelligence

Radiology departments today are faced with many challenges to improve operational efficiency, performance, and quality. Many organizations rely on antiquated, paper-based methods to review their historical performance and understand their operations. With increased workloads, geographically dispersed image acquisition and reading sites, and rapidly changing technologies, this approach is increasingly untenable. A Web-based dashboard was constructed to automate the extraction, processing, and display of indicators and thereby provide useful and current data for twice-monthly departmental operational meetings. The feasibility of extracting specific metrics from clinical information systems was evaluated as part of a longer-term effort to build a radiology business intelligence architecture. Operational data were extracted from clinical information systems and stored in a centralized data warehouse. Higher-level analytics were performed on the centralized data, a process that generated indicators in a dynamic Web-based graphical environment that proved valuable in discussion and root cause analysis. Results aggregated over a 24-month period since implementation suggest that this operational business intelligence reporting system has provided significant data for driving more effective management decisions to improve productivity, performance, and quality of service in the department.