The Experts below are selected from a list of 15327 Experts worldwide ranked by ideXlab platform
Riccardo Bellazzi - One of the best experts on this subject based on the ideXlab platform.
-
EMBC - Improving risk-stratification of Diabetes complications using Temporal Data Mining
2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015Co-Authors: Lucia Sacchi, Arianna Dagliati, Daniele Segagni, Paola Leporati, Luca Chiovato, Riccardo BellazziAbstract:To understand which factor trigger worsened disease control is a crucial step in Type 2 Diabetes (T2D) patient management. The MOSAIC project, funded by the European Commission under the FP7 program, has been designed to integrate heterogeneous Data sources and provide decision support in chronic T2D management through patients' continuous stratification. In this work we show how Temporal Data Mining can be fruitfully exploited to improve risk stratification. In particular, we exploit administrative Data on drug purchases to divide patients in meaningful groups. The detection of drug consumption patterns allows stratifying the population on the basis of subjects' purchasing attitude. Merging these findings with clinical values indicates the relevance of the applied methods while showing significant differences in the identified groups. This extensive approach emphasized the exploitation of administrative Data to identify patterns able to explain clinical conditions.
-
Analyzing complex patients' Temporal histories: new frontiers in Temporal Data Mining.
Methods in Molecular Biology, 2014Co-Authors: Lucia Sacchi, Arianna Dagliati, Riccardo BellazziAbstract:In recent years, Data coming from hospital information systems (HIS) and local healthcare organizations have started to be intensively used for research purposes. This rising amount of available Data allows reconstructing the compete histories of the patients, which have a strong Temporal component. This chapter introduces the major challenges faced by Temporal Data Mining researchers in an era when huge quantities of complex clinical Temporal Data are becoming available. The analysis is focused on the peculiar features of this kind of Data and describes the methodological and technological aspects that allow managing such complex framework. The chapter shows how heterogeneous Data can be processed to derive a homogeneous representation. Starting from this representation, it illustrates different techniques for jointly analyze such kind of Data. Finally, the technological strategies that allow creating a common Data warehouse to gather Data coming from different sources and with different formats are presented.
-
Temporal Data Mining and process Mining techniques to identify cardiovascular risk-associated clinical pathways in Type 2 diabetes patients
IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), 2014Co-Authors: Arianna Dagliati, Lucia Sacchi, Paola Leporati, Luca Chiovato, Carlo Cerra, Pasquale De Cata, John. H. Holmes, Riccardo BellazziAbstract:In this work we present the results of a workflow Mining approach to analyze complex Temporal Datasets of Type 2 Diabetes (T2D) patients. The research has been performed within the EU project MOSAIC, which gathers T2D patients' Data coming from three European hospitals and a local health care agency. The main idea underlying our approach is to use a suite of Temporal Data Mining methods in order to derive healthcare pathways. The approach starts by processing raw Data, derived from heterogeneous Data sources, and create event logs, which contain meaningful healthcare activities. Once event logs have been obtained and tasks and transitions defined, it is possible to explore how state-of-art process Mining techniques can be used to gain insights into T2D patients care. In the experimental section of this paper we illustrate the results of this approach applied to an integrated Data repository containing clinical and administrative Data of 1,020 T2D patients.
-
AMIA - Temporal Data Mining for the assessment of the costs related to diabetes mellitus pharmacological treatment.
AMIA ... Annual Symposium proceedings. AMIA Symposium, 2009Co-Authors: Stefano Concaro, Lucia Sacchi, Carlo Cerra, Mario Stefanelli, Pietro Fratino, Riccardo BellazziAbstract:Diabetes care and chronic disease management represent Data-intensive contexts which allow Local Healthcare Agencies (ASL) to collect a huge amount of information. Time is often an essential component of such information, given the strong importance of the Temporal evolution of the considered disease and of its treatment. In this paper we show the application of a Temporal Data Mining technique to extract Temporal association rules over an integrated repository including both administrative and clinical Data related to a sample of diabetic patients. We will show how the method can be used to highlight cases and conditions which lead to the highest pharmaceutical costs. Considering the perspective of a Regional Healthcare Agency, this method could be properly exploited to assess the overall standards and quality of care, while lowering costs.
-
AIME - Temporal Data Mining of HIV Registries: Results from a 25 Years Follow-Up
Artificial Intelligence in Medicine, 2009Co-Authors: Paloma Chausa, Riccardo Bellazzi, Lucia Sacchi, César Cáceres, Agathe León, Felipe García, Enrique J. GómezAbstract:The Human Immunodeficiency Virus (HIV) causes a pandemic infection in humans, with millions of people infected every year. Although the Highly Active Antiretroviral Therapy reduced the number of AIDS cases since 1996 by significantly increasing the disease-free survival time, the therapy failure rate is still high due to HIV treatment complexity. To better understand the changes in the outcomes of HIV patients we have applied Temporal Data Mining techniques to the analysis of the Data collected since 1981 by the Infectious Diseases Unit of the Hospital Clinic in Barcelona, Spain. We run a precedence Temporal rule extraction algorithm on three different Temporal periods, looking for two types of treatment failures: viral failure and toxic failure, corresponding to events of clinical interest to assess the treatment outcomes. The analysis allowed to extract different typical patterns related to each period and to meaningfully interpret the previous and current behaviour of HIV therapy.
Niall Rooney - One of the best experts on this subject based on the ideXlab platform.
-
Temporal Data Mining for smart homes
Lecture Notes in Computer Science, 2006Co-Authors: Mykola Galushka, Dave Patterson, Niall RooneyAbstract:Temporal Data Mining is a relatively new area of research in computer science. It can provide a large variety of different methods and techniques for handling and analyzing Temporal Data generated by smart-home environments. Temporal Data Mining in general fits into a two level architecture, where initially a transformation technique reduces Data dimensionality in the first level and indexing techniques provide efficient access to the Data in the second level. This infrastructure of Temporal Data Mining provides the basis for high-level Data Mining operations such as clustering, classification, rule discovery and prediction. These operations can form the basis for developing different smart-home applications, capable of addressing a number of situations occurring within this environment. This paper outlines the main Temporal Data Mining techniques available and provides examples of where they can be applied within a smart home environment.
-
Designing Smart Homes - Temporal Data Mining for smart homes
Lecture Notes in Computer Science, 2006Co-Authors: Mykola Galushka, Dave Patterson, Niall RooneyAbstract:Temporal Data Mining is a relatively new area of research in computer science. It can provide a large variety of different methods and techniques for handling and analyzing Temporal Data generated by smart-home environments. Temporal Data Mining in general fits into a two level architecture, where initially a transformation technique reduces Data dimensionality in the first level and indexing techniques provide efficient access to the Data in the second level. This infrastructure of Temporal Data Mining provides the basis for high-level Data Mining operations such as clustering, classification, rule discovery and prediction. These operations can form the basis for developing different smart-home applications, capable of addressing a number of situations occurring within this environment. This paper outlines the main Temporal Data Mining techniques available and provides examples of where they can be applied within a smart home environment.
Mykola Galushka - One of the best experts on this subject based on the ideXlab platform.
-
Temporal Data Mining for smart homes
Lecture Notes in Computer Science, 2006Co-Authors: Mykola Galushka, Dave Patterson, Niall RooneyAbstract:Temporal Data Mining is a relatively new area of research in computer science. It can provide a large variety of different methods and techniques for handling and analyzing Temporal Data generated by smart-home environments. Temporal Data Mining in general fits into a two level architecture, where initially a transformation technique reduces Data dimensionality in the first level and indexing techniques provide efficient access to the Data in the second level. This infrastructure of Temporal Data Mining provides the basis for high-level Data Mining operations such as clustering, classification, rule discovery and prediction. These operations can form the basis for developing different smart-home applications, capable of addressing a number of situations occurring within this environment. This paper outlines the main Temporal Data Mining techniques available and provides examples of where they can be applied within a smart home environment.
-
Designing Smart Homes - Temporal Data Mining for smart homes
Lecture Notes in Computer Science, 2006Co-Authors: Mykola Galushka, Dave Patterson, Niall RooneyAbstract:Temporal Data Mining is a relatively new area of research in computer science. It can provide a large variety of different methods and techniques for handling and analyzing Temporal Data generated by smart-home environments. Temporal Data Mining in general fits into a two level architecture, where initially a transformation technique reduces Data dimensionality in the first level and indexing techniques provide efficient access to the Data in the second level. This infrastructure of Temporal Data Mining provides the basis for high-level Data Mining operations such as clustering, classification, rule discovery and prediction. These operations can form the basis for developing different smart-home applications, capable of addressing a number of situations occurring within this environment. This paper outlines the main Temporal Data Mining techniques available and provides examples of where they can be applied within a smart home environment.
Naren Ramakrishnan - One of the best experts on this subject based on the ideXlab platform.
-
Temporal Data Mining Approaches for Sustainable Chiller Management in Data Centers
ACM Trans. Intell. Syst. Technol., 2011Co-Authors: Debprakash Patnaik, Manish Marwah, Naren RamakrishnanAbstract:Practically every large IT organization hosts Data centers---a mix of computing elements, storage systems, networking, power, and cooling infrastructure---operated either in-house or outsourced to major vendors. A significant element of modern Data centers is their cooling infrastructure, whose efficient and sustainable operation is a key ingredient to the “always-on” capability of Data centers. We describe the design and implementation of CAMAS (Chiller Advisory and MAnagement System), a Temporal Data Mining solution to mine and manage chiller installations. CAMAS embodies a set of algorithms for processing multivariate time-series Data and characterizes sustainability measures of the patterns mined. We demonstrate three key ingredients of CAMAS---motif Mining, association analysis, and dynamic Bayesian network inference---that help bridge the gap between low-level, raw, sensor streams, and the high-level operating regions and features needed for an operator to efficiently manage the Data center. The effectiveness of CAMAS is demonstrated by its application to a real-life production Data center managed by HP.
-
Temporal Data Mining for Neuroscience
GPU Computing Gems Emerald Edition, 2011Co-Authors: Wu-chun Feng, Debprakash Patnaik, Yong Cao, Naren RamakrishnanAbstract:Publisher Summary This chapter presents a solution that uses graphics processing units (GPUs) to mine spike train Datasets. Specifically, the solution delivers a novel mapping of a “finite state machine for Data Mining” onto the GPU while simultaneously addressing a wide range of neuronal input characteristics. This solution ultimately transforms the task of Temporal Data Mining of spike trains from a batch-oriented process towards a real-time one. Multielectrode arrays (MEAs) capture neuronal spike streams in real time, thus providing dynamic perspectives into brain function. Mining such spike streams from these MEAs is critical toward understanding the firing patterns of neurons and gaining insight into the underlying cellular activity. However, the acquisition rate of neuronal Data places a tremendous computational burden on the subsequent Temporal Data Mining of these spike streams. An MEA records spiking action potentials from an ensemble of neurons, and after various preprocessing steps, these neurons yield a spike train Dataset that provides a real-time dynamic perspective into brain function. Key problems of interest include identifying sequences of firing neurons, deterMining their characteristic delays, and reconstructing the functional connectivity of neuronal circuits. Addressing these problems can provide critical insights into the cellular activity recorded in the neuronal tissue. For the first time, neuroscientists can enjoy the benefits of Data Mining algorithms without needing access to costly and specialized clusters of workstations.
-
Accelerator-Oriented Algorithm Transformation for Temporal Data Mining
arXiv: Distributed Parallel and Cluster Computing, 2009Co-Authors: Debprakash Patnaik, Yong Cao, Sean P. Ponce, Naren RamakrishnanAbstract:Temporal Data Mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train Data. While application scientists have been able to readily gather multi-neuronal Datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia's GTX 280 offers a cost-effective option to bring these capabilities to the neuroscientist's desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting "in-the-large" issues such as problem decomposition as well as "in-the-small" issues such as Data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly Data-parallel mapping strategies. Applications to many Datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
-
Sustainable Operation and Management of Data Center Chillers Using Temporal Data Mining
Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, 2009Co-Authors: Debprakash Patnaik, Manish Marwah, Ratnesh Sharma, Naren RamakrishnanAbstract:Motivation: Data centers are a critical component of modern IT infrastructure but are also among the worst environmental offenders through their increasing energy usage and the resulting large carbon footprints. Efficient management of Data centers, including power management, networking, and cooling infrastructure, is hence crucial to sustainability. In the absence of a 'first-principles' approach to manage these complex components and their interactions, Data-driven approaches have become attractive and tenable. Results: We present a Temporal Data Mining solution to model and optimize performance of Data center chillers, a key component of the cooling infrastructure. It helps bridge raw, numeric, time-series information from sensor streams toward higher level characterizations of chiller behavior, suitable for a Data center engineer. To aid in this transduction, Temporal Data streams are first encoded into a symbolic representation, next run-length encoded segments are mined to form frequent motifs in time series, and finally these metrics are evaluated by their contributions to sustainability. A key innovation in our application is the ability to intersperse "don't care" transitions (e.g., transients) in continuous-valued time series Data, an advantage we inherit by the application of frequent episode Mining to symbolized representations of numeric time series. Our approach provides both qualitative and quantitative characterizations of the sensor streams to the Data center engineer, to aid him in tuning chiller operating characteristics. This system is currently being prototyped for a Data center managed by HP and experimental results from this application reveal the promise of our approach.
-
NPC - Accelerator-Oriented Algorithm Transformation for Temporal Data Mining
2009 Sixth IFIP International Conference on Network and Parallel Computing, 2009Co-Authors: Debprakash Patnaik, Yong Cao, Sean P. Ponce, Naren RamakrishnanAbstract:Temporal Data Mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train Data. While application scientists have been able to readily gather multi-neuronal Datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia's GTX 280 offers a cost-effective option to bring these capabilities to the neuroscientist's desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting 'in-the-large' issues such as problem decomposition as well as 'in-the-small' issues such as Data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly Data-parallel mapping strategies. Applications to many Datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
Debprakash Patnaik - One of the best experts on this subject based on the ideXlab platform.
-
Temporal Data Mining Approaches for Sustainable Chiller Management in Data Centers
ACM Trans. Intell. Syst. Technol., 2011Co-Authors: Debprakash Patnaik, Manish Marwah, Naren RamakrishnanAbstract:Practically every large IT organization hosts Data centers---a mix of computing elements, storage systems, networking, power, and cooling infrastructure---operated either in-house or outsourced to major vendors. A significant element of modern Data centers is their cooling infrastructure, whose efficient and sustainable operation is a key ingredient to the “always-on” capability of Data centers. We describe the design and implementation of CAMAS (Chiller Advisory and MAnagement System), a Temporal Data Mining solution to mine and manage chiller installations. CAMAS embodies a set of algorithms for processing multivariate time-series Data and characterizes sustainability measures of the patterns mined. We demonstrate three key ingredients of CAMAS---motif Mining, association analysis, and dynamic Bayesian network inference---that help bridge the gap between low-level, raw, sensor streams, and the high-level operating regions and features needed for an operator to efficiently manage the Data center. The effectiveness of CAMAS is demonstrated by its application to a real-life production Data center managed by HP.
-
Temporal Data Mining for Neuroscience
GPU Computing Gems Emerald Edition, 2011Co-Authors: Wu-chun Feng, Debprakash Patnaik, Yong Cao, Naren RamakrishnanAbstract:Publisher Summary This chapter presents a solution that uses graphics processing units (GPUs) to mine spike train Datasets. Specifically, the solution delivers a novel mapping of a “finite state machine for Data Mining” onto the GPU while simultaneously addressing a wide range of neuronal input characteristics. This solution ultimately transforms the task of Temporal Data Mining of spike trains from a batch-oriented process towards a real-time one. Multielectrode arrays (MEAs) capture neuronal spike streams in real time, thus providing dynamic perspectives into brain function. Mining such spike streams from these MEAs is critical toward understanding the firing patterns of neurons and gaining insight into the underlying cellular activity. However, the acquisition rate of neuronal Data places a tremendous computational burden on the subsequent Temporal Data Mining of these spike streams. An MEA records spiking action potentials from an ensemble of neurons, and after various preprocessing steps, these neurons yield a spike train Dataset that provides a real-time dynamic perspective into brain function. Key problems of interest include identifying sequences of firing neurons, deterMining their characteristic delays, and reconstructing the functional connectivity of neuronal circuits. Addressing these problems can provide critical insights into the cellular activity recorded in the neuronal tissue. For the first time, neuroscientists can enjoy the benefits of Data Mining algorithms without needing access to costly and specialized clusters of workstations.
-
Accelerator-Oriented Algorithm Transformation for Temporal Data Mining
arXiv: Distributed Parallel and Cluster Computing, 2009Co-Authors: Debprakash Patnaik, Yong Cao, Sean P. Ponce, Naren RamakrishnanAbstract:Temporal Data Mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train Data. While application scientists have been able to readily gather multi-neuronal Datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia's GTX 280 offers a cost-effective option to bring these capabilities to the neuroscientist's desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting "in-the-large" issues such as problem decomposition as well as "in-the-small" issues such as Data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly Data-parallel mapping strategies. Applications to many Datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
-
Sustainable Operation and Management of Data Center Chillers Using Temporal Data Mining
Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, 2009Co-Authors: Debprakash Patnaik, Manish Marwah, Ratnesh Sharma, Naren RamakrishnanAbstract:Motivation: Data centers are a critical component of modern IT infrastructure but are also among the worst environmental offenders through their increasing energy usage and the resulting large carbon footprints. Efficient management of Data centers, including power management, networking, and cooling infrastructure, is hence crucial to sustainability. In the absence of a 'first-principles' approach to manage these complex components and their interactions, Data-driven approaches have become attractive and tenable. Results: We present a Temporal Data Mining solution to model and optimize performance of Data center chillers, a key component of the cooling infrastructure. It helps bridge raw, numeric, time-series information from sensor streams toward higher level characterizations of chiller behavior, suitable for a Data center engineer. To aid in this transduction, Temporal Data streams are first encoded into a symbolic representation, next run-length encoded segments are mined to form frequent motifs in time series, and finally these metrics are evaluated by their contributions to sustainability. A key innovation in our application is the ability to intersperse "don't care" transitions (e.g., transients) in continuous-valued time series Data, an advantage we inherit by the application of frequent episode Mining to symbolized representations of numeric time series. Our approach provides both qualitative and quantitative characterizations of the sensor streams to the Data center engineer, to aid him in tuning chiller operating characteristics. This system is currently being prototyped for a Data center managed by HP and experimental results from this application reveal the promise of our approach.
-
NPC - Accelerator-Oriented Algorithm Transformation for Temporal Data Mining
2009 Sixth IFIP International Conference on Network and Parallel Computing, 2009Co-Authors: Debprakash Patnaik, Yong Cao, Sean P. Ponce, Naren RamakrishnanAbstract:Temporal Data Mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train Data. While application scientists have been able to readily gather multi-neuronal Datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia's GTX 280 offers a cost-effective option to bring these capabilities to the neuroscientist's desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting 'in-the-large' issues such as problem decomposition as well as 'in-the-small' issues such as Data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly Data-parallel mapping strategies. Applications to many Datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
