0.25
0.5
0.75
1.25
1.5
1.75
2
Improving the reproducibility of experiments and reusability of research outputs in complex data analysis
Published on 2019-06-2868 Views
The advances in science are heavily based on the premise of the concept of a trusted discovery, provided that the performed research is done correctly, and reproducible by other scientists. In order t
Related categories
Presentation
Improving the reproducibility of experiments and reusability of research outputs in complex data analysis00:00
Outline01:34
Reproducibility in Science 02:45
Reproducibility in Artificial Intelligence (AI)04:55
Reusable research outputs and FAIR06:53
FAIR data principles09:55
Adding semantics to research outputs: Ontologies12:18
Semantic web technologies14:37
Complex data analysis18:23
Example task: Structured output prediction20:40
How can we improve the reproducibility and reusability (1)23:58
How can we improve the reproducibility and reusability (2)27:04
How can we improve the reproducibility and reusability (3)30:02
Storing experiments, research outputs and their annotations32:01
Relevant ontological representations for machine learning and data mining34:47
OntoDM : a set of modular ontologies for data mining36:17
Generic ontology of datatypes OntoDT37:02
Examples of different datatypes in data mining37:47
Example of dataset taxonomy constructed using OntoDT40:08
Ontology of core data mining entities OntoDM core41:14
OntoDM core structure41:45
Representation of complex analysis tasks43:42
How do we represent complex tasks?43:47
OntoDM KDD44:04
OntoDM KDD structure44:25
IMPERATRIX project44:52
Project objectives46:45
Vision for the IMPERATRIX prototype system47:55
Current work48:57
Prototype system50:09
Outline of the system prototype50:36
Summary50:50