Search for: All records

Abstract Continuous monitoring of blood glucose (BG) levels is a key aspect of diabetes management. Patients with Type-1 diabetes (T1D) require an effective tool to monitor these levels in order to make appropriate decisions regarding insulin administration and food intake to keep BG levels in target range. Effectively and accurately predicting future BG levels at multi-time steps ahead benefits a patient with diabetes by helping them decrease the risks of extremes in BG including hypo- and hyperglycemia. In this study, we present a novel multi-component deep learning model that predicts the BG levels in a multi-step look ahead fashion. The model is evaluated both quantitatively and qualitatively on actual blood glucose data for 97 patients. For the prediction horizon (PH) of 30 mins, the average values forroot mean squared error(RMSE),mean absolute error(MAE),mean absolute percentage error(MAPE), andnormalized mean squared error(NRMSE) are$$23.22 \pm 6.39$$ $23.22\pm 6.39$mg/dL, 16.77 ± 4.87 mg/dL,$$12.84 \pm 3.68$$ $12.84\pm 3.68$and$$0.08 \pm 0.01$$ $0.08\pm 0.01$respectively. When Clarke and Parkes error grid analyses were performed comparing predicted BG with actual BG, the results showed average percentage of points in Zone A of$$80.17 \pm 9.20$$ $80.17\pm 9.20$and$$84.81 \pm 6.11,$$ $84.81\pm 6.11,$respectively. We offer this tool as a mechanism to enhance the predictive capabilities of algorithms for patients with T1D. 

Scientific data, generated by computational models or from experiments, are typically results of nonlinear interactions among several latent processes. Such datasets are typically high-dimensional and exhibit strong temporal correlations. Better understanding of the underlying processes requires mapping such data to a low-dimensional manifold where the dynamics of the latent processes are evident. While nonlinear spectral dimensionality reduction methods, e.g., Isomap, and their scalable variants, are conceptually fit candidates for obtaining such a mapping, the presence of the strong temporal correlation in the data can significantly impact these methods. In this paper, we first show why such methods fail when dealing with dynamic process data. A novel method, Entropy-Isomap, is proposed to handle this shortcoming. We demonstrate the effectiveness of the proposed method in the context of understanding the fabrication process of organic materials. The resulting low-dimensional representation correctly characterizes the process control variables and allows for informative visualization of the material morphology evolution. 

Analyzing database access logs is a key part of performance tuning, intrusion detection, benchmark development, and many other database administration tasks. Unfortunately, it is common for production databases to deal with millions or more queries each day, so these logs must be summarized before they can be used. Designing an appropriate summary encoding requires trading off between conciseness and information content. For example: simple workload sampling may miss rare, but high impact queries. In this paper, we present LogR, a lossy log compression scheme suitable for use in many automated log analytics tools, as well as for human inspection. We formalize and analyze the space/fidelity trade-off in the context of a broader family of “pattern” and “pattern mixture” log encodings to which LogR belongs. We show through a series of experiments that LogR compressed encodings can be created efficiently, come with provable information-theoretic bounds on their accuracy, and outperform state-of-art log summarization strategies. 

Manifold learning based methods have been widely used for non-linear dimensionality reduction (NLDR). However, in many practical settings, the need to process streaming data is a challenge for such methods, owing to the high computational complexity involved. Moreover, most methods operate under the assumption that the input data is sampled from a single manifold, embedded in a high dimensional space. We propose a method for streaming NLDR when the observed data is either sampled from multiple manifolds or irregularly sampled from a single manifold. We show that existing NLDR methods, such as Isomap, fail in such situations, primarily because they rely on smoothness and continuity of the underlying manifold, which is violated in the scenarios explored in this paper. However, the proposed algorithm is able to learn effectively in presence of multiple, and potentially intersecting, manifolds, while allowing for the input data to arrive as a massive stream. 

Database access logs are the starting point for many forms of database administration, from database performance tuning, to security auditing, to benchmark design, and many more. Unfortunately, query logs are also large and unwieldy, and it can be difficult for an analyst to extract broad patterns from the set of queries found therein. Clustering is a natural first step towards understanding the massive query logs. However, many clustering methods rely on the notion of pairwise similarity, which is challenging to compute for SQL queries, especially when the underlying data and database schema is unavailable. We investigate the problem of computing similarity between queries, relying only on the query structure. We conduct a rigorous evaluation of three query similarity heuristics proposed in the literature applied to query clustering on multiple query log datasets, representing different types of query workloads. To improve the accuracy of the three heuristics, we propose a generic feature engineering strategy, using classical query rewrites to standardize query structure. The proposed strategy results in a significant improvement in the performance of all three similarity heuristics. 

Insider threats to databases in the financial sector have become a very serious and pervasive security problem. This paper proposes a framework to analyze access patterns to databases by clustering SQL queries issued to the database. Our system Ettu works by grouping queries with other similarly structured queries. The small number of intent groups that result can then be efficiently labeled by human operators. We show how our system is designed and how the components of the system work. Our preliminary results show that our system accurately models user intent.