false
OasisLMS
Login
Catalog
Deep Dive: An Introduction to AI in Critical Care ...
Data Issues Upstream of Machine Learning
Data Issues Upstream of Machine Learning
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Video Summary
Omar Bedouin, Chief of the Division of Data Sciences at the Telemedicine Advanced Technology Research Center, discusses the importance of understanding data provenance in research. Highlighting his extensive experience, he emphasizes awareness of data's origin and potential limitations. Bedouin explains the concept of competence, emphasizing that in research, it’s crucial to recognize we often don't know what we don't know, which can lead to faulty assumptions. He illustrates scenarios in research where data definitions can be misleading, underscoring the necessity of thorough data investigation and not taking definitions at face value. He discusses the concept of data provenance as a documentation trail and its role in validating research data. Despite this, Bedouin warns of common issues such as measurement error, network drops, replication errors, and reference data errors, advocating for practices like using validated data sets, stating assumptions, and employing exploratory data analysis to enhance data reliability and mitigate issues.
Meta Tag
Concept
Data Provenance
Concept
Data Origin
Concept
Measurement Error
Concept
Missing Data
Concept
Data Quality
Keywords
data provenance
research validation
data reliability
exploratory data analysis
measurement error
Data Provenance
Data Origin
Measurement Error
Missing Data
Data Quality
×
Please select your language
1
English