en
0.25
0.5
0.75
1.25
1.5
1.75
2
Machine Learning with Knowledge Graphs
Published on Jul 30, 201414148 Views
Most successful applications of statistical machine learning focus on response learning or signal-reaction learning where an output is produced as a direct response to an input. An important feature i
Related categories
Chapter list
Machine Learning with Knowledge Graphs00:00
Prelude00:16
Prelude01:22
Overview01:31
What is Machine Learning?01:34
Classification02:34
Typical Classifiers02:57
Deep Learning Neural Networks04:44
Detecting Cats in Images05:46
Where from here?06:15
Challenges06:50
Vision07:26
γλαῦκας εἰς Ἀθήνας κομίζειν07:43
Requirement: Understanding of the World08:05
Biomedical Ontologies08:14
For the First Time there Exist Sizable General Ontologies: DBpedia, YAGO, Freebase, Knowledge Graph08:37
Linked Open Data (Semantic Web)09:11
Triple Graphs09:15
Knowledge Bases are Triple Graphs09:24
Overview10:09
Canonical Relational Machine Learning Task10:20
II. Relational Learning with Latent Features11:40
With Latent Features We Get Collective Learning15:50
Model with Polynomial Basis Functions17:25
Mapping to a Tensor Factorization Problem18:04
RESCAL Factorization19:26
Cost Functions20:38
Iterative Update21:02
RESCAL for Different -arities21:31
Scalabilty21:49
Leading Performance in Link prediction on benchmark data sets23:07
Cora Data: Entity Resolution24:42
Overview26:06
Yago2 Core Ontology26:11
Classification: Type Prediction26:32
Writer‘s Nationality: Demonstrating Collective Learning27:58
Learning a Taxonomy (-> Ontology)28:58
Extensions: Nonnegative RESCAL29:27
Extensions: Proofs and Bounds30:44
Overview31:23
Machine Learning with Structured Data and Ontologies31:32
Clinical Data Intelligence33:37
Predicting Diagnoses and Procedures40:53
Machine Learning with Images and Ontologies41:45
References and Related Work42:02
Conclusions43:26
Thank you45:27