Learning from heterogeneous temporal data in electronic health records

Zhao, Jing; Papapetrou, Panagiotis; Asker, Lars; Boström, Henrik

doi:10.1016/j.jbi.2016.11.006

Cited by 77 publications

(53 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, a discrete, symbolic representation of the time series is obtained, which may be handled by traditional pattern mining algorithms [17]. Due to its simplicity, many studies have pursued such an approach, for instance [37,38]; still, its downsides typically include information loss, and a reduced interpretability of the final results. Other strategies make use of patterns that may be extracted straight from the numerical data, such as time series shapelets.…”

Section: Time Series Shapeletsmentioning

confidence: 99%

J48SS: A Novel Decision Tree Approach for the Handling of Sequential and Time Series Data

Brunello

Marzano²,

Montanari

et al. 2019

Computers

View full text Add to dashboard Cite

Temporal information plays a very important role in many analysis tasks, and can be encoded in at least two different ways. It can be modeled by discrete sequences of events as, for example, in the business intelligence domain, with the aim of tracking the evolution of customer behaviors over time. Alternatively, it can be represented by time series, as in the stock market to characterize price histories. In some analysis tasks, temporal information is complemented by other kinds of data, which may be represented by static attributes, e.g., categorical or numerical ones. This paper presents J48SS, a novel decision tree inducer capable of natively mixing static (i.e., numerical and categorical), sequential, and time series data for classification purposes. The novel algorithm is based on the popular C4.5 decision tree learner, and it relies on the concepts of frequent pattern extraction and time series shapelet generation. The algorithm is evaluated on a text classification task in a real business setting, as well as on a selection of public UCR time series datasets. Results show that it is capable of providing competitive classification performances, while generating highly interpretable models and effectively reducing the data preparation effort.

show abstract

Section: Time Series Shapeletsmentioning

confidence: 99%

J48SS: A Novel Decision Tree Approach for the Handling of Sequential and Time Series Data

Brunello

Marzano²,

Montanari

et al. 2019

Computers

View full text Add to dashboard Cite

show abstract

“…1 of [8]. We have already selected our data formation clustering method to be Random Dynamic Subsequent method proposed in [27]. Further when performing machine learning heuristics we consider the proposed deep patient representation using unsupervised deep feature learning method shown in Fig.…”

Section: Evaluating Proposed Solutionmentioning

confidence: 99%

Context Aware SmartHealth Cloud Platform for Medical Diagnostics

Shafqat¹,

Abbasi²,

Khan³

et al. 2018

ijacsa

View full text Add to dashboard Cite

Abstract-Healthcare has seen a great evolution in current era in terms of new computer technologies. Intensive medical data is generated that opens up research in healthcare analytics. Coping with this intensive data along with making it meaningful to deliver knowledge and be able to make decisions are the most important tasks. To deduce the authenticity of the data on basis of precision, correction, associations and true meaning is important to validate the understanding of correct semantics. In case of medical diagnosis to form accurate understanding of associations while removing ambiguity and forming a correct picture of the case is of utmost importance. To come up with the right metrics for the diagnostic solution we have explored the known criteria to validate healthcare analytics techniques involved in formation of diagnosis that results in betterment and safety of patients under observations and heading towards possible treatments. In this work, we have proposed a thematic taxonomy for the comparison of existing healthcare analytics techniques with emphasis on diabetes and its underlying diseases. This analysis lead us to propose a data model for hybrid distributed simulation model for future Context Aware SmartHealth cloud platform for diagnostics. This platform is designed to inherit smartness of unsupervised learning which in turn would keep updating itself under supervised learning by qualified experts. Finally, the accuracy would be determined using HUM approach with biomarkers or a better accuracy model than AUC. The recommended action plan is also presented.

show abstract

“…Rather, there exist temporal relations connecting them, representing the variable nature of an individual's health. These relations cannot be described by one feature or a single value, but require longitudinal observations with a series of values over time [66].…”

Section: Temporal Relationsmentioning

confidence: 99%

Beyond Volume: The Impact of Complex Healthcare Data on the Machine Learning Pipeline

Feldman

Faust

et al. 2017

Towards Integrative Machine Learning and Knowledge Extraction

View full text Add to dashboard Cite

Abstract. From medical charts to national census, healthcare has traditionally operated under a paper-based paradigm. However, the past decade has marked a long and arduous transformation bringing healthcare into the digital age. Ranging from electronic health records, to digitized imaging and laboratory reports, to public health datasets, today, healthcare now generates an incredible amount of digital information. Such a wealth of data presents an exciting opportunity for integrated machine learning solutions to address problems across multiple facets of healthcare practice and administration. Unfortunately, the ability to derive accurate and informative insights requires more than the ability to execute machine learning models. Rather, a deeper understanding of the data on which the models are run is imperative for their success. While a significant effort has been undertaken to develop models able to process the volume of data obtained during the analysis of millions of digitalized patient records, it is important to remember that volume represents only one aspect of the data. In fact, drawing on data from an increasingly diverse set of sources, healthcare data presents an incredibly complex set of attributes that must be accounted for throughout the machine learning pipeline. This chapter focuses on highlighting such challenges, and is broken down into three distinct components, each representing a phase of the pipeline. We begin with attributes of the data accounted for during preprocessing, then move to considerations during model building, and end with challenges to the interpretation of model output. For each component, we present a discussion around data as it relates to the healthcare domain and offer insight into the challenges each may impose on the efficiency of machine learning techniques.

show abstract

Learning from heterogeneous temporal data in electronic health records

Cited by 77 publications

References 66 publications

J48SS: A Novel Decision Tree Approach for the Handling of Sequential and Time Series Data

J48SS: A Novel Decision Tree Approach for the Handling of Sequential and Time Series Data

Context Aware SmartHealth Cloud Platform for Medical Diagnostics

Beyond Volume: The Impact of Complex Healthcare Data on the Machine Learning Pipeline

Contact Info

Product

Resources

About