2012 11th International Conference on Machine Learning and Applications 2012
DOI: 10.1109/icmla.2012.41
|View full text |Cite
|
Sign up to set email alerts
|

Semantic Data Types in Machine Learning from Healthcare Data

Abstract: Healthcare is particularly rich in semantic information and background knowledge describing data. This paper discusses a number of semantic data types that can be found in healthcare data, presents how the semantics can be extracted from existing sources including the Unified Medical Language System (UMLS), discusses how the semantics can be used in both supervised and unsupervised learning, and presents an example rule learning system that implements several of these types. Results from three example applicat… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
3
0

Year Published

2013
2013
2021
2021

Publication Types

Select...
3
2
1

Relationship

1
5

Authors

Journals

citations
Cited by 8 publications
(5 citation statements)
references
References 13 publications
0
3
0
Order By: Relevance
“…), graph (values are linked together by edges in a graph, i.e., ), set (multiple values can be selected at the same time, i.e., [diagnosis={diabetes, hypertension, obesity}]), set-defined (similar to set, but with additional structure added on top of values, i.e., diagnoses in previous example form a hierarchy), interval (numeric values with defined addition and subtraction, zero is not defined), ratio (numeric values with defined multiplication and division), cyclic-ratio (numeric values forming a cycle, i.e., [angle=276..15]), and absolute (numeric values with only order defines and no operations permitted, i.e., social security number). These attribute types along with additional examples are explained in [38]. Attribute types are critical when attempting to generalize and reason with rules, as well as apply constructive induction methods, i.e.…”
Section: Aq21 Semantics and Ontologiesmentioning
confidence: 99%
“…), graph (values are linked together by edges in a graph, i.e., ), set (multiple values can be selected at the same time, i.e., [diagnosis={diabetes, hypertension, obesity}]), set-defined (similar to set, but with additional structure added on top of values, i.e., diagnoses in previous example form a hierarchy), interval (numeric values with defined addition and subtraction, zero is not defined), ratio (numeric values with defined multiplication and division), cyclic-ratio (numeric values forming a cycle, i.e., [angle=276..15]), and absolute (numeric values with only order defines and no operations permitted, i.e., social security number). These attribute types along with additional examples are explained in [38]. Attribute types are critical when attempting to generalize and reason with rules, as well as apply constructive induction methods, i.e.…”
Section: Aq21 Semantics and Ontologiesmentioning
confidence: 99%
“…Jednakże, pojęcie danych medycznych ma znacznie szersze znaczenie. Można wyróżnić pewne określone rodzaje danych medycznych pod względem semantycznym, między innymi [13]:  kategorie, operujące na zbiorach nieuporządkowanych, na przykład grupa krwi {0+, A+, B+ …},  kategorie uporządkowane, na przykład {niski, średni, wysoki},  cykliczne, dla których zbiór wartości zawiera powtarzający się, skończony zbiór wartości, na przykład minuty, godziny, dni tygodnia,  hierarchiczne, operujące na usystematyzowanych właściwościach, na przykład oznaczenia jednostek chorobowych w kategoriach Międzynarodowej Klasyfikacji Chorób ICD-9 [22],  grafy, zawierające elementy i relacje między nimi. Przykładem jest opis anatomiczny ludzkiego ciała lub lokalizacja ogniska zapalnego,  zbiory danych, często będące podzbiorami skończonego zbioru zawierającego wszystkie zestawy wartości (skończony zbiór diagnoz).…”
Section: Dane Medyczneunclassified
“…Але поняття медичних даних, має набагато ширше значення. Можна виділити деякі специфічні види медичних даних, між іншим, з семантичної точки зору [13]: 69 agregacji danych np. danej populacji, z danej kliniki lub pacjentów cierpiących na jedno schorzenie.…”
Section: медичні даніunclassified
“…The system stops further prediction so as to avoid propagation of errors.If the previous data as well as the current data is valid and then the current value is an abnormality, then an abnormality is reported. Otherwise, backup of the data is taken and regression function (4) computes the prediction of the next data.…”
Section: Predictionmentioning
confidence: 99%
“…The models devised are used to extrapolate on the information available so that a future abnormality can be predicted. Machine learning approaches [3] [4] have also been aided by other computational developments to reduce the errors arising out of the models [5].…”
Section: Introductionmentioning
confidence: 99%