SciTePress - Affiliation Profile

Research.Publish.Connect.

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

Name:
Country:
Subject:

Advanced Search Affiliations Search

If you're looking for an exact phrase use quotation marks on text fields.

Proceedings

Proceedings Search

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

Papers

Papers Search

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

Authors

Authors Search

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

Advanced Search

Affiliation

Select a Subject
Refine Ontology Subject

KNOWLEDGE DISCOVERY AND DATA MINING RESEARCH GROUP, United States

Context (please use complete words only)

Field

Sort by

19 Papers

Early Prediction of MRSA Infections using Electronic Health Records

Topics: Data Mining; Physiological Modeling; Electronic Health Records and Standards; Pattern Recognition and Machine Learning; Medical Informatics

Authors: Thomas Hartvigsen ¹ , Cansu Sen ¹ , Sarah Brownell ² , Erin Teeple ¹ , Xiangnan Kong ¹ and Elke Rundensteiner ¹

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² Simmons College, United States

Digital Object Identifier: 10.5220/0006599601560167

Proceedings of HEALTHINF 2018, ISBN: 978-989-758-281-3

Quick Abstract

Despite eradication efforts, Methicillin-resistant Staphylococcus aureus (MRSA) remains a common cause of serious hospital-acquired infections (HAI) in the United States. Electronic Health Record (EHR) systems capture MRSA infection events along with detailed patient information preceding diagnosis. In this work, we design and apply machine learning methods to support early recognition of MRSA infection by estimating risk at several time points during hospitalization. We use EHR data including on-admission and throughout-stay patient information. On-admission features capture clinical and non-clinical information while throughout-stay features include vital signs, medications, laboratory studies, and other clinical assessments. We evaluate prediction accuracy achieved by core Machine Learning methods, namely Logistic Regression, Support Vector Machine, and Random Forest classifiers, when mining these different types of EHR features to detect patterns predictive of MRSA infection. We evaluate classification performance using MIMIC III – a critical care data set comprised of 12 years of patient records from the Beth Israel Deaconess Medical Center Intensive Care Unit in Boston, MA. Our methods can achieve near-perfect MRSA prediction accuracies one day before documented clinical diagnosis. Also, they perform well for early MRSA prediction many days in advance of diagnosis. These findings underscore the potential clinical applicability of machine learning techniques.

Related Ontology Subjects/Areas/Topics (12)

Artificial Intelligence	Electronic Health Records and Standards	Physiological Modeling
Biomedical Engineering	Enterprise Information Systems	Sensor Networks
Data Mining	Health Information Systems	Signal Processing
Databases and Information Systems Integration	Pattern Recognition and Machine Learning	Soft Computing

One Size Does Not Fit All: An Ensemble Approach Towards Information Extraction from Adverse Drug Event Narratives

Topics: Data Mining; Pattern Recognition and Machine Learning; Medical Informatics

Authors: Susmitha Wunnava ¹ , Xiao Qin ¹ , Tabassum Kakar ¹ , Xiangnan Kong ¹ , Elke A. Rundensteiner ¹ , Sanjay K. Sahoo ² and Suranjan De ²

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² U.S. Food and Drug Administration, United States

Digital Object Identifier: 10.5220/0006600201760188

Proceedings of HEALTHINF 2018, ISBN: 978-989-758-281-3

Quick Abstract

Recognizing named entities in Adverse Drug Reactions narratives is a fundamental step towards extracting valuable patient information from unstructured text into a structured thus actionable format. This then unlocks advanced data analytics towards intelligent pharmacovigilance. Yet existing biomedical named entity recognition (NER) tools are limited in their ability to identify certain entity types from these domain-specific narratives and result in significant performance differences in terms of accuracy. To address these challenges, we propose an ensemble approach that integrates a rich variety of named entity recognizers to procure the final result. First, one critical problem faced by NER in the biomedical context is that the data is highly skewed. That is, only 1% of words belong to a certain medical entity type, such as, the reason for medication usage compared to all other non-reason words. We propose a balanced, under-sampled bagging strategy that is dependent on the imbalance level to overcome the class imbalance problem. Second, we present an ensemble of heterogeneous recognizers approach that leverages a novel ensemble combiner. Our experimental results show that for biomedical text datasets: (i) a balanced learning environment along with an Ensemble of Heterogeneous Classifiers constantly improves the performance over individual base learners and, (ii) stacking-based ensemble combiner methods outperform simple Majority Voting by 0.30 F-measure.

Related Ontology Subjects/Areas/Topics (10)

Artificial Intelligence	Enterprise Information Systems	Sensor Networks
Biomedical Engineering	Health Information Systems	Signal Processing
Data Mining	Pattern Recognition and Machine Learning	Soft Computing
Databases and Information Systems Integration

Predicting Outcome of Ischemic Stroke Patients using Bootstrap Aggregating with M5 Model Trees

Topics: Therapeutic Systems and Technologies; Development of Assistive Technology; Software Systems in Medicine; Medical Informatics

Authors: Ahmedul Kabir ¹ , Carolina Ruiz ¹ , Sergio A. Alvarez ² and Majaz Moonis ³

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² Boston College, United States , ³ Univ. Massachusetts Medical School, United States

Digital Object Identifier: 10.5220/0006282001780187

Proceedings of HEALTHINF 2017/BIOSTEC, ISBN: 978-989-758-213-4

Quick Abstract

The objective of our study is to predict the clinical outcome of ischemic stroke patients after 90 days of stroke using the modified Rankin Scale (mRS) score. After experimentation with various regression techniques, we discovered that using M5 model trees to predict the score and then using bootstrap aggregating as a meta-learning technique produces the best prediction results. The same regression when followed by classification also performs better than regular multi-class classification. In this paper, we present the methodology used, and compare the results with other standard predictive techniques. We also analyze the results to provide insights on the factors that affect stroke outcomes.

Related Ontology Subjects/Areas/Topics (5)

Biomedical Engineering		Health Information Systems		Therapeutic Systems and Technologies
Development of Assistive Technology		Software Systems in Medicine

Deviation-based Dynamic Time Warping for Clustering Human Sleep

Topics: Pattern Recognition; Medical Signal Acquisition, Analysis and Processing; Physiological Processes and Bio-Signal Modeling, Non-Linear Dynamics

Authors: Chiying Wang ¹ , Sergio A. Alvarez ² , Carolina Ruiz ¹ and Majaz Moonis ³

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² Boston College, United States , ³ U. of Massachusetts Medical School, United States

Digital Object Identifier: 10.5220/0005729200880095

Proceedings of BIOSIGNALS 2016/BIOSTEC, ISBN: 978-989-758-170-0

Quick Abstract

In this paper, we propose two versions of a modified dynamic time warping approach for comparing discrete time series. This approach is motivated by the observation that the distribution of dynamic time warping paths between pairs of human sleep time series is concentrated around the path of constant slope. Both versions use a penalty term for the deviation between the warping path and the path of constant slope for a given pair of time series. In the first version, global weighted dynamic time warping, the penalty term is added as a post-processing step after a standard dynamic time warping computation, yielding a modified similarity metric that can be used for time series clustering. The second version, stepwise deviation-based dynamic time warping, incorporates the penalty term into the dynamic programming optimization itself, yielding modified optimal warping paths, together with a similarity metric. Clustering experiments over synthetic data, as well as over human sleep data, show that the proposed methods yield significantly improved accuracy and generative log likelihood as compared with standard dynamic time warping.

Related Ontology Subjects/Areas/Topics (17)

Applications and Services	Health Engineering and Technology Applications	Pattern Recognition
Artificial Intelligence	Human-Computer Interaction	Physiological Computing Systems
Biomedical Engineering	Medical Image Detection, Acquisition, Analysis and Processing	Physiological Processes and Bio-Signal Modeling, Non-Linear Dynamics
Biomedical Signal Processing	Methodologies and Methods	Sensor Networks
Computer Vision, Visualization and Computer Graphics	Neurocomputing	Soft Computing
Data Manipulation	Neurotechnology, Electronics and Informatics

Model-based Clustering of Ischemic Stroke Patients

Topics: Datamining; Medical and Nursing Informatics; Pattern Recognition and Machine Learning

Authors: Ahmedul Kabir ¹ , Carolina Ruiz ¹ , Sergio Alvarez ² , Nazish Riaz ³ and Majaz Moonis ³

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² Boston College, United States , ³ University of Massachusetts Medical School, United States

Digital Object Identifier: 10.5220/0005278101720181

Proceedings of HEALTHINF 2015/BIOSTEC, ISBN: 978-989-758-068-0

Quick Abstract

The objective of our study is to find meaningful groups in the data of ischemic stroke patients using unsupervised clustering. The data are modeled using Gaussian mixture models with a variety of covariance structures. Cluster parameters in each of these models are estimated by maximum likelihood via the Expectation-Maximization algorithm. The best models are then selected by relying on information-theoretic criteria. It is observed that the stroke patients can be grouped into a small number of medically relevant clusters that are defined primarily by the presence of diabetes and atrial fibrillation. Characteristics of the clusters found are discussed, using statistical comparisons and data visualization.

Related Ontology Subjects/Areas/Topics (17)

Artificial Intelligence	Data Mining	Medical and Nursing Informatics
Biomedical Engineering	Databases and Information Systems Integration	Pattern Recognition and Machine Learning
Business Analytics	Datamining	Sensor Networks
Cardiovascular Technologies	Enterprise Information Systems	Signal Processing
Computing and Telecommunications in Cardiology	Health Engineering and Technology Applications	Soft Computing
Data Engineering	Health Information Systems

A Novel Feature Generation Method for Sequence Classification - Mutated Subsequence Generation

Topics: Sequence Analysis; Data Mining and Machine Learning; Algorithms and Software Tools; Pattern Recognition, Clustering and Classification

Authors: Hao Wan , Carolina Ruiz and Joseph Beck

Affiliation: Worcester Polytechnic Institute, United States

Digital Object Identifier: 10.5220/0004808200680079

Proceedings of BIOINFORMATICS 2014/BIOSTEC, ISBN: 978-989-758-012-3

Quick Abstract

In this paper, we present a new feature generation algorithm for sequence data sets called Mutated Subsequence Generation (MSG). Given a data set of sequences, the MSG algorithm generates features from these sequences by incorporating mutative positions in subsequences. We compare this algorithm with other sequence-based feature generation algorithms, including position-based, k-grams, and k-gapped pairs. Our experiments show that the MSG algorithm outperforms these other algorithms in domains in which presence, not specific location, of sequential patterns discriminate among classes in a data set.

Related Ontology Subjects/Areas/Topics (6)

Algorithms and Software Tools		Biomedical Engineering		Pattern Recognition, Clustering and Classification
Bioinformatics		Data Mining and Machine Learning		Sequence Analysis

Complex Patten Processing in Spatio-temporal Databases

Topics: Data Management for Analytics; Query Processing and Optimization; Nosql Databases

Authors: Yang Zheng , Annies Ductan , Devin Thomas and Mohamed Y. Eltabakh

Affiliation: Worcester Polytechnic Institute (WPI), United States

Digital Object Identifier: 10.5220/0004992401570169

Proceedings of DATA 2014, ISBN: 978-989-758-035-2

Quick Abstract

The increasing complexity of spatio-temporal applications has caused the underlying queries to be more sophisticated and usually carry complex semantics. As a result, the traditional spatio-temporal query types, e.g., range, kNN, and aggregation queries, have become just building blocks in more complex query plans. In this paper, we present the STEPQ system, which is an extensible spatio-temporal query engine for complex pattern processing over spatio-temporal data. STEPQ enables full-fledged and optimized integration between spatiotemporal queries and complex event processing (CEP). This integration enables expressing complex queries that execute the desired application semantics without the need for indifferent middle-aware or application level support. The system is implemented using TerraLib module on top of PostgreSQL DBMSs. The experimental evaluation demonstrates the feasibility and practicality of the STEPQ system, and the efficiency of the proposed optimizations.

Related Ontology Subjects/Areas/Topics (6)

Data Engineering		Data Management for Analytics		Nosql Databases
Data Management and Quality		Databases and Data Security		Query Processing and Optimization

FunctionGuard - A Query Engine for Expensive Scientific Functions in Relational Databases

Topics: Data Structures and Data Management Algorithms; Query Processing and Optimization

Authors: Anh Pham and Mohamed Eltabakh

Affiliation: Worcester Polytechnic Institute (WPI), United States

Digital Object Identifier: 10.5220/0004992300950106

Proceedings of DATA 2014, ISBN: 978-989-758-035-2

Quick Abstract

Expensive user-defined functions impose unique challenges to database management systems at query time. This is mostly due to the black-box nature of these functions, the in-ability to optimize their internals, and the potential inefficiency of the common optimization heuristics, e.g., “selection-push-down’. Moreover, the in- creasing diversity of modern scientific applications that depend on DBMSs and, at the same time, extensively use expensive UDFs is mandating the design and development of efficient techniques to support these expensive functions. In this paper, we propose the “FunctionGuard” system that leverages disk-based persistent caching in novel ways to achieve across-queries optimizations for expensive UDFs. The unique features of FunctionGuard include: (1) Dynamic extraction of dependencies between the UDFs and the data sources and identifying the potential cacheable functions, (2) Cache-aware query optimization through newly introduced query operators, (3) Proactive cache refreshing that partially migrates the cost of the expensive calls from the query time to the idle and under-utilized times, and (4) Integration with the state-of-art techniques that generate efficient query plans under the presence of expensive functions. The system is implemented within PostgreSQL DBMS, and the results show the effectiveness of the proposed algorithms and optimizations.

Related Ontology Subjects/Areas/Topics (5)

Data Engineering		Data Structures and Data Management Algorithms		Query Processing and Optimization
Data Management and Quality		Databases and Data Security

Semi-Markov Modeling-Clustering of Human Sleep with Efficient Initialization and Stopping

Topics: Pattern Recognition; Medical Signal Acquisition, Analysis and Processing; Physiological Processes and Bio-Signal Modeling, Non-Linear Dynamics

Authors: Chiying Wang ¹ , Sergio A. Alvarez ² , Carolina Ruiz ¹ and Majaz Moonis ³

Affiliations: ¹ Worcester Polytechnic Institute, United States , ² Boston College, United States , ³ University of Massachusetts Medical School, United States

Digital Object Identifier: 10.5220/0004824900610068

Proceedings of BIOSIGNALS 2014/BIOSTEC, ISBN: 978-989-758-011-6

Quick Abstract

Collective Dynamical Modeling-Clustering (CDMC) is an algorithmic framework for time series dynamical modeling and clustering using probabilistic state-transition models. In this paper, an efficient initialization technique based on Itakura slope-constrained Dynamic Time Warping is applied to CDMC. Semi-Markov chains are used as the dynamical models. Experimental evaluation demonstrates the effectiveness of the proposed approach in providing more realistic dynamical modeling of sleep stage dynamics than Markov models, with improved clustering quality and convergence speed as compared with pseudorandom initialization.

Related Ontology Subjects/Areas/Topics (17)

Applications and Services	Health Engineering and Technology Applications	Pattern Recognition
Artificial Intelligence	Human-Computer Interaction	Physiological Computing Systems
Biomedical Engineering	Medical Image Detection, Acquisition, Analysis and Processing	Physiological Processes and Bio-Signal Modeling, Non-Linear Dynamics
Biomedical Signal Processing	Methodologies and Methods	Sensor Networks
Computer Vision, Visualization and Computer Graphics	Neurocomputing	Soft Computing
Data Manipulation	Neurotechnology, Electronics and Informatics

The Values on Academic Frontier-based Approach’ Implementation

Topics: Collaborative Learning; Community Building

Authors: Nuo Liu ¹ , Shengli Yang ² and Ziyong Liu ³

Affiliations: ¹ University of Electronic Science and Technology of China, China , ² Zhejiang University of Technology, China , ³ Worcester Polytechnic Institute, United States

Digital Object Identifier: 10.5220/0004978901350139

Proceedings of CSEDU 2014, ISBN: 978-989-758-022-2

Quick Abstract

Academic Frontier-based Approach (AFA) is a student-centred instructional approach used to promote active and deep learning by involving students in investigating academic frontier issues in a collaborative environment. In this paper, we discuss the student surveys conducted at the end of the study period to solicit feedback from students on their learning experience. It is suggested that AFA takes the students as the center, consolidates the academic knowledge of the students. Through the teamwork and autonomous learning, students know more about the frontiers of science and technology and open their horizon. In addition, the ability of language organization, speech, compressive resistance, communication and cooperation are all increased. Furthermore, students’ interest in studying is increased because of the autonomous learning. It is expected the popularization of the consecutive courses will certainly improve these important abilities further more. And it is hopeful that students will take advantage from the sustainable ability of autonomous learning in their future career life.

Related Ontology Subjects/Areas/Topics (4)