SlideShare a Scribd company logo
1 of 37
Systems that learn and reason
Frank van Harmelen
Knowledge Representation & Reasoning Group
Vrije Universiteit Amsterdam
Creative Commons License
CC BY 3.0:
Allowed to copy, redistribute
remix & transform
But must attribute
1
Direction of travel in modern AI
“making artificial intelligence more human-
centred”
AI systems should “support people and be
competent partners”
But:
“current AI systems are often incompetent
because
they lack background and contextual
knowledge” 2
So what’s holding AI back?
For a long time,
AI researchers have locked themselves
in one of two towers 3
“current AI systems are often incompetent because
they lack background and contextual knowledge”
and “they cannot explain their actions”
AI: a tale of
two towers?
Statistical AI Symbolic AI
AI: two religions?
But our brain uses both …
6
In our brain (roughly)
Data
Driven
Knowledge
Driven
Symbolic
Knowledge
Logic
Reasoning
Connectionist
Data
Statistics
Learning
The AI pendulum!
8
So let’s compare
9
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
10
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
Very large knowledge graphs:
Billions of facts and rules,
But expensive to build & maintain:
- Crowd sourcing
11
Strengths & Weaknesses
10M training samples
Symbolic Connectionist
Construction Human effort Data hunger
Scalable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
4.8M training games
12
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
worse with
more data
worse with
less data
13
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
Strengths & Weaknesses
Symbolic Statistical
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
15
“black box problem”
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
quality

generality  16
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
17
Class: 793
Label: n04209133 (shower cap)
Certainty: 99.7%
Problem:
“out of sample generalizability”
Strengths & Weaknesses
Symbolic Connectionist
Construction Human effort Data hunger
Scaleable +/- +/-
Explainable + -
Generalisable Performance cliff Performance cliff
18
Can we get them
to collaborate?
19
Learning to add facts to knowledge graph
(“knowledge graph completion”)
Rolling
Stones
Angi
Beat It
Jackson
Publish
20
Angi
Rolling
Stones
Publish
From: Predict:
Yesterday
Publish
Beatles
…..
Publish
…..
prediction
algorithm
21
Learning to add facts to a knowledge graph
(“knowledge graph completion”)
Example: Inductive Logic Programming,
rule mining
(anyBURL on knowledge graphs)
22
parent(Mary,Vicky).
parent(Mary,Andre).
parent(Carrey,Vickey).
mother(Mary,Vicky).
mother(Mary,Andy).
father(Carrey,Vickey).
father(Carrey,Andy).
parent(X,Y) :- mother(X,Y).
parent(X,Y) :- father(X,Y)
parent(Carrey,Andy).
Inductive Logic Programming:
application
23
Labels
Facts Theory
Active
Inactive
Informed ML using KR
queen
crown
wears
24
shower
cap
?
Predict Select
crown?
showercap?
Explainable ML using KR
queen
wears
25
shower
cap
?
Predict Explain
crown?
Context aware ML using KR
(informed ML)
See survey of 100+ systems in Von Rueden et al, Learning, 2019
flower?
cushion?
“Parts of a chair are:
cushion and armrest”
“Given the context of chair,
a cushion is much more likely
than a flower”
P(cushion|chair) >> P(flower|chair)
Learning intermediate abstractions
:- see( , 3), see( , 5), add(3,5,8).
End-to-end:
2x 784
inputs
19 outputs
8
Learning intermediate abstractions
29
Neural
Back
end
Symbolic
Front
end
Learning intermediate abstractions
Faster adaptation to changes;
Better transfer learning
Learning intermediate abstractions
Faster adaptation to changes;
Better transfer learning
Injecting knowledge into rule-learning
32
33
0
0
Injecting knowledge into rule-learning
• Learn the rules 𝑟𝑑and linear terms l()
• Learn the weights 𝛼 and 𝛽
Injecting knowledge into rule-learning
• Learn the rules 𝑟𝑑and linear terms l()
• Experts give the rules 𝑟𝑐 and 𝑟𝑐
• Learn the weights 𝛼 and 𝛽
Concluding
remarks
36
1. Machine Learning can benefit from injecting
symbolic knowledge (knowledge graphs)
• Explanations
• Ranking hypotheses
• Transfer learning
• Sample efficiency
• Pipeline engineering
2. The required symbolic knowledge is available
at very large scale + corresponding tools
“Why learn what you already know?”
37

More Related Content

What's hot

RDBMS vs Hadoop vs Spark
RDBMS vs Hadoop vs SparkRDBMS vs Hadoop vs Spark
RDBMS vs Hadoop vs SparkLaxmi8
 
An overview of data warehousing and OLAP technology
An overview of  data warehousing and OLAP technology An overview of  data warehousing and OLAP technology
An overview of data warehousing and OLAP technology Nikhatfatima16
 
Δομή Βιογραφικού Σημειώματος
Δομή Βιογραφικού ΣημειώματοςΔομή Βιογραφικού Σημειώματος
Δομή Βιογραφικού ΣημειώματοςNickos Nickolopoulos
 
Τα οφέλη της τηλεόρασης, Εργασία μαθητή
Τα οφέλη της τηλεόρασης, Εργασία μαθητήΤα οφέλη της τηλεόρασης, Εργασία μαθητή
Τα οφέλη της τηλεόρασης, Εργασία μαθητήΣΟΦΙΑ ΦΕΛΛΑΧΙΔΟΥ
 
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdf
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdfΠαιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdf
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdfΣάσα Καραγιαννίδου - Πέννα
 
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"Eva Skanavi
 
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!""Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"Rozana Papakrasa
 
What is new in Apache Hive 3.0?
What is new in Apache Hive 3.0?What is new in Apache Hive 3.0?
What is new in Apache Hive 3.0?DataWorks Summit
 
ποιήματα για το νερό
ποιήματα για το νερόποιήματα για το νερό
ποιήματα για το νερόorfayannis
 
Ο όρκος στην Ιλιάδα και την αρχαιότητα
Ο όρκος στην Ιλιάδα και την αρχαιότηταΟ όρκος στην Ιλιάδα και την αρχαιότητα
Ο όρκος στην Ιλιάδα και την αρχαιότηταΣύλια Ζέττα-Silia Zetta
 
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣ
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣΗ ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣ
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣd tampouris
 
Big Data can be fun!
Big Data can be fun!Big Data can be fun!
Big Data can be fun!Bruno Aziza
 
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...ΕΥΗ ΚΑΡΟΥΝΙΑ
 
Παιδική Λογοτεχνία και Φιλαναγνωστικές Δραστηριότητες
Παιδική Λογοτεχνία και Φιλαναγνωστικές ΔραστηριότητεςΠαιδική Λογοτεχνία και Φιλαναγνωστικές Δραστηριότητες
Παιδική Λογοτεχνία και Φιλαναγνωστικές ΔραστηριότητεςΧρύσα Κουράκη
 
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσης
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσηςδιδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσης
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσηςMaria Georgoutsou
 
συνδετικες – διαρθρωτικες λεξεις και εκφρασεις για τη συνδεση προτασε...
συνδετικες  –  διαρθρωτικες  λεξεις  και  εκφρασεις  για  τη συνδεση  προτασε...συνδετικες  –  διαρθρωτικες  λεξεις  και  εκφρασεις  για  τη συνδεση  προτασε...
συνδετικες – διαρθρωτικες λεξεις και εκφρασεις για τη συνδεση προτασε...Eleni Kots
 

What's hot (20)

RDBMS vs Hadoop vs Spark
RDBMS vs Hadoop vs SparkRDBMS vs Hadoop vs Spark
RDBMS vs Hadoop vs Spark
 
An overview of data warehousing and OLAP technology
An overview of  data warehousing and OLAP technology An overview of  data warehousing and OLAP technology
An overview of data warehousing and OLAP technology
 
Spark etl
Spark etlSpark etl
Spark etl
 
Δομή Βιογραφικού Σημειώματος
Δομή Βιογραφικού ΣημειώματοςΔομή Βιογραφικού Σημειώματος
Δομή Βιογραφικού Σημειώματος
 
Τα οφέλη της τηλεόρασης, Εργασία μαθητή
Τα οφέλη της τηλεόρασης, Εργασία μαθητήΤα οφέλη της τηλεόρασης, Εργασία μαθητή
Τα οφέλη της τηλεόρασης, Εργασία μαθητή
 
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdf
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdfΠαιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdf
Παιγχνιώδεις δραστηριότητες με θέμα τους πλανήτες(1).pdf
 
Glossa
GlossaGlossa
Glossa
 
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"
ΠΟΛΙΤΙΣΤΙΚΟ ΠΡΟΓΡΑΜΜΑ "ΔΙΚΑΙΩΜΑΤΑ ΤΟΥ ΠΑΙΔΙΟΥ & ΔΙΑΦΟΡΕΤΙΚΟΤΗΤΑ"
 
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!""Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"
"Τρέφομαι υγιεινά...σώμα και δόντια να έχω γερά!!!"
 
What is new in Apache Hive 3.0?
What is new in Apache Hive 3.0?What is new in Apache Hive 3.0?
What is new in Apache Hive 3.0?
 
ποιήματα για το νερό
ποιήματα για το νερόποιήματα για το νερό
ποιήματα για το νερό
 
Data science
Data scienceData science
Data science
 
Ο όρκος στην Ιλιάδα και την αρχαιότητα
Ο όρκος στην Ιλιάδα και την αρχαιότηταΟ όρκος στην Ιλιάδα και την αρχαιότητα
Ο όρκος στην Ιλιάδα και την αρχαιότητα
 
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣ
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣΗ ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣ
Η ΓΛΩΣΣΑ του ΣΩΜΑΤΟΣ
 
Big Data can be fun!
Big Data can be fun!Big Data can be fun!
Big Data can be fun!
 
οι τέσσερις εποχές.pdf
οι τέσσερις εποχές.pdfοι τέσσερις εποχές.pdf
οι τέσσερις εποχές.pdf
 
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...
ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ΤΟΥ ΕΙΜΙ,ΟΡΙΣΤΙΚΗ ΕΝΕΣΤΩΤΑ ,ΜΕΛΛΟΝΤΑ ΒΑΡΥΤΟΝΩΝ ΡΗΜΑΤΩΝ ΕΝΕΡ...
 
Παιδική Λογοτεχνία και Φιλαναγνωστικές Δραστηριότητες
Παιδική Λογοτεχνία και Φιλαναγνωστικές ΔραστηριότητεςΠαιδική Λογοτεχνία και Φιλαναγνωστικές Δραστηριότητες
Παιδική Λογοτεχνία και Φιλαναγνωστικές Δραστηριότητες
 
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσης
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσηςδιδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσης
διδασκαλία εννοιών πληροφορικής στο νηπιαγωγείο μια μελέτη περίπτωσης
 
συνδετικες – διαρθρωτικες λεξεις και εκφρασεις για τη συνδεση προτασε...
συνδετικες  –  διαρθρωτικες  λεξεις  και  εκφρασεις  για  τη συνδεση  προτασε...συνδετικες  –  διαρθρωτικες  λεξεις  και  εκφρασεις  για  τη συνδεση  προτασε...
συνδετικες – διαρθρωτικες λεξεις και εκφρασεις για τη συνδεση προτασε...
 

Similar to Systems that learn and reason | Frank Van Harmelen

Modular design patterns for systems that learn and reason: a boxology
Modular design patterns for systems that learn and reason: a boxologyModular design patterns for systems that learn and reason: a boxology
Modular design patterns for systems that learn and reason: a boxologyFrank van Harmelen
 
Introduction to the Artificial Intelligence and Computer Vision revolution
Introduction to the Artificial Intelligence and Computer Vision revolutionIntroduction to the Artificial Intelligence and Computer Vision revolution
Introduction to the Artificial Intelligence and Computer Vision revolutionDarian Frajberg
 
A primer on Artificial Intelligence (AI) and Machine Learning (ML)
A primer on Artificial Intelligence (AI) and Machine Learning (ML)A primer on Artificial Intelligence (AI) and Machine Learning (ML)
A primer on Artificial Intelligence (AI) and Machine Learning (ML)Yacine Ghalim
 
Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine LearningArtificial Intelligence and Machine Learning
Artificial Intelligence and Machine LearningMykola Dobrochynskyy
 
Big, Open, Data and Semantics for Real-World Application Near You
Big, Open, Data and Semantics for Real-World Application Near YouBig, Open, Data and Semantics for Real-World Application Near You
Big, Open, Data and Semantics for Real-World Application Near YouBiplav Srivastava
 
Y conf talk - Andrej Karpathy
Y conf talk - Andrej KarpathyY conf talk - Andrej Karpathy
Y conf talk - Andrej KarpathySze Siong Teo
 
Information Visualization for Knowledge Discovery: An Introduction
Information Visualization for Knowledge Discovery: An IntroductionInformation Visualization for Knowledge Discovery: An Introduction
Information Visualization for Knowledge Discovery: An IntroductionKrist Wongsuphasawat
 
AI and the Future of Work [TUG-CO, 11/15/23]
AI and the Future of Work [TUG-CO, 11/15/23]AI and the Future of Work [TUG-CO, 11/15/23]
AI and the Future of Work [TUG-CO, 11/15/23]Matt Small
 
AI – Risks, Opportunities and Ethical Issues April 2023.pdf
AI – Risks, Opportunities and Ethical Issues April 2023.pdfAI – Risks, Opportunities and Ethical Issues April 2023.pdf
AI – Risks, Opportunities and Ethical Issues April 2023.pdfAdam Ford
 
Palo alto university rotary club talk Sep 29, 2107
Palo alto university rotary club talk Sep 29, 2107Palo alto university rotary club talk Sep 29, 2107
Palo alto university rotary club talk Sep 29, 2107Charles Martin
 
DWX 2018 Session about Artificial Intelligence, Machine and Deep Learning
DWX 2018 Session about Artificial Intelligence, Machine and Deep LearningDWX 2018 Session about Artificial Intelligence, Machine and Deep Learning
DWX 2018 Session about Artificial Intelligence, Machine and Deep LearningMykola Dobrochynskyy
 
Cognitive systems institute talk 8 june 2017 - v.1.0
Cognitive systems institute talk   8 june 2017 - v.1.0Cognitive systems institute talk   8 june 2017 - v.1.0
Cognitive systems institute talk 8 june 2017 - v.1.0diannepatricia
 
From Turing To Humanoid Robots - Ramón López de Mántaras
From Turing To Humanoid Robots - Ramón López de MántarasFrom Turing To Humanoid Robots - Ramón López de Mántaras
From Turing To Humanoid Robots - Ramón López de MántarasMachine Learning Valencia
 
AI Introduction: AI is the new electricity (by Slash)
AI Introduction: AI is the new electricity (by Slash)AI Introduction: AI is the new electricity (by Slash)
AI Introduction: AI is the new electricity (by Slash)Andries De Vos
 
The 2030 Workplace - Evolving Your Organization with Emerging Tech
The 2030 Workplace - Evolving Your Organization with Emerging TechThe 2030 Workplace - Evolving Your Organization with Emerging Tech
The 2030 Workplace - Evolving Your Organization with Emerging TechSage Franch
 
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...NETWAYS
 
DL Classe 0 - You can do it
DL Classe 0 - You can do itDL Classe 0 - You can do it
DL Classe 0 - You can do itGregory Renard
 
Deep Learning Class #0 - You Can Do It
Deep Learning Class #0 - You Can Do ItDeep Learning Class #0 - You Can Do It
Deep Learning Class #0 - You Can Do ItHolberton School
 

Similar to Systems that learn and reason | Frank Van Harmelen (20)

Modular design patterns for systems that learn and reason: a boxology
Modular design patterns for systems that learn and reason: a boxologyModular design patterns for systems that learn and reason: a boxology
Modular design patterns for systems that learn and reason: a boxology
 
Introduction to the Artificial Intelligence and Computer Vision revolution
Introduction to the Artificial Intelligence and Computer Vision revolutionIntroduction to the Artificial Intelligence and Computer Vision revolution
Introduction to the Artificial Intelligence and Computer Vision revolution
 
A primer on Artificial Intelligence (AI) and Machine Learning (ML)
A primer on Artificial Intelligence (AI) and Machine Learning (ML)A primer on Artificial Intelligence (AI) and Machine Learning (ML)
A primer on Artificial Intelligence (AI) and Machine Learning (ML)
 
Artificial Intelligence and Machine Learning
Artificial Intelligence and Machine LearningArtificial Intelligence and Machine Learning
Artificial Intelligence and Machine Learning
 
Big, Open, Data and Semantics for Real-World Application Near You
Big, Open, Data and Semantics for Real-World Application Near YouBig, Open, Data and Semantics for Real-World Application Near You
Big, Open, Data and Semantics for Real-World Application Near You
 
Where Does It Break?
Where Does It Break?Where Does It Break?
Where Does It Break?
 
Y conf talk - Andrej Karpathy
Y conf talk - Andrej KarpathyY conf talk - Andrej Karpathy
Y conf talk - Andrej Karpathy
 
Information Visualization for Knowledge Discovery: An Introduction
Information Visualization for Knowledge Discovery: An IntroductionInformation Visualization for Knowledge Discovery: An Introduction
Information Visualization for Knowledge Discovery: An Introduction
 
AI and the Future of Work [TUG-CO, 11/15/23]
AI and the Future of Work [TUG-CO, 11/15/23]AI and the Future of Work [TUG-CO, 11/15/23]
AI and the Future of Work [TUG-CO, 11/15/23]
 
AI – Risks, Opportunities and Ethical Issues April 2023.pdf
AI – Risks, Opportunities and Ethical Issues April 2023.pdfAI – Risks, Opportunities and Ethical Issues April 2023.pdf
AI – Risks, Opportunities and Ethical Issues April 2023.pdf
 
Palo alto university rotary club talk Sep 29, 2107
Palo alto university rotary club talk Sep 29, 2107Palo alto university rotary club talk Sep 29, 2107
Palo alto university rotary club talk Sep 29, 2107
 
DWX 2018 Session about Artificial Intelligence, Machine and Deep Learning
DWX 2018 Session about Artificial Intelligence, Machine and Deep LearningDWX 2018 Session about Artificial Intelligence, Machine and Deep Learning
DWX 2018 Session about Artificial Intelligence, Machine and Deep Learning
 
Cognitive systems institute talk 8 june 2017 - v.1.0
Cognitive systems institute talk   8 june 2017 - v.1.0Cognitive systems institute talk   8 june 2017 - v.1.0
Cognitive systems institute talk 8 june 2017 - v.1.0
 
From Turing To Humanoid Robots - Ramón López de Mántaras
From Turing To Humanoid Robots - Ramón López de MántarasFrom Turing To Humanoid Robots - Ramón López de Mántaras
From Turing To Humanoid Robots - Ramón López de Mántaras
 
20181212 ibm aot
20181212 ibm aot20181212 ibm aot
20181212 ibm aot
 
AI Introduction: AI is the new electricity (by Slash)
AI Introduction: AI is the new electricity (by Slash)AI Introduction: AI is the new electricity (by Slash)
AI Introduction: AI is the new electricity (by Slash)
 
The 2030 Workplace - Evolving Your Organization with Emerging Tech
The 2030 Workplace - Evolving Your Organization with Emerging TechThe 2030 Workplace - Evolving Your Organization with Emerging Tech
The 2030 Workplace - Evolving Your Organization with Emerging Tech
 
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...
stackconf 2020 | Artificial Intelligence? – more like Artificial Stupidity! b...
 
DL Classe 0 - You can do it
DL Classe 0 - You can do itDL Classe 0 - You can do it
DL Classe 0 - You can do it
 
Deep Learning Class #0 - You Can Do It
Deep Learning Class #0 - You Can Do ItDeep Learning Class #0 - You Can Do It
Deep Learning Class #0 - You Can Do It
 

More from Connected Data World

Graph Abstractions Matter by Ora Lassila
Graph Abstractions Matter by Ora LassilaGraph Abstractions Matter by Ora Lassila
Graph Abstractions Matter by Ora LassilaConnected Data World
 
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...Connected Data World
 
How to get started with Graph Machine Learning
How to get started with Graph Machine LearningHow to get started with Graph Machine Learning
How to get started with Graph Machine LearningConnected Data World
 
The years of the graph: The future of the future is here
The years of the graph: The future of the future is hereThe years of the graph: The future of the future is here
The years of the graph: The future of the future is hereConnected Data World
 
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2Connected Data World
 
From Taxonomies and Schemas to Knowledge Graphs: Part 3
From Taxonomies and Schemas to Knowledge Graphs: Part 3From Taxonomies and Schemas to Knowledge Graphs: Part 3
From Taxonomies and Schemas to Knowledge Graphs: Part 3Connected Data World
 
In Search of the Universal Data Model
In Search of the Universal Data ModelIn Search of the Universal Data Model
In Search of the Universal Data ModelConnected Data World
 
Graph in Apache Cassandra. The World’s Most Scalable Graph Database
Graph in Apache Cassandra. The World’s Most Scalable Graph DatabaseGraph in Apache Cassandra. The World’s Most Scalable Graph Database
Graph in Apache Cassandra. The World’s Most Scalable Graph DatabaseConnected Data World
 
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...Connected Data World
 
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...Connected Data World
 
Semantic similarity for faster Knowledge Graph delivery at scale
Semantic similarity for faster Knowledge Graph delivery at scaleSemantic similarity for faster Knowledge Graph delivery at scale
Semantic similarity for faster Knowledge Graph delivery at scaleConnected Data World
 
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...Connected Data World
 
Schema, Google & The Future of the Web
Schema, Google & The Future of the WebSchema, Google & The Future of the Web
Schema, Google & The Future of the WebConnected Data World
 
RAPIDS cuGraph – Accelerating all your Graph needs
RAPIDS cuGraph – Accelerating all your Graph needsRAPIDS cuGraph – Accelerating all your Graph needs
RAPIDS cuGraph – Accelerating all your Graph needsConnected Data World
 
Elegant and Scalable Code Querying with Code Property Graphs
Elegant and Scalable Code Querying with Code Property GraphsElegant and Scalable Code Querying with Code Property Graphs
Elegant and Scalable Code Querying with Code Property GraphsConnected Data World
 
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...Connected Data World
 
Graph for Good: Empowering your NGO
Graph for Good: Empowering your NGOGraph for Good: Empowering your NGO
Graph for Good: Empowering your NGOConnected Data World
 
What are we Talking About, When we Talk About Ontology?
What are we Talking About, When we Talk About Ontology?What are we Talking About, When we Talk About Ontology?
What are we Talking About, When we Talk About Ontology?Connected Data World
 

More from Connected Data World (20)

Graph Abstractions Matter by Ora Lassila
Graph Abstractions Matter by Ora LassilaGraph Abstractions Matter by Ora Lassila
Graph Abstractions Matter by Ora Lassila
 
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...
Κnowledge Architecture: Combining Strategy, Data Science and Information Arch...
 
How to get started with Graph Machine Learning
How to get started with Graph Machine LearningHow to get started with Graph Machine Learning
How to get started with Graph Machine Learning
 
Graphs in sustainable finance
Graphs in sustainable financeGraphs in sustainable finance
Graphs in sustainable finance
 
The years of the graph: The future of the future is here
The years of the graph: The future of the future is hereThe years of the graph: The future of the future is here
The years of the graph: The future of the future is here
 
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2
From Taxonomies and Schemas to Knowledge Graphs: Parts 1 & 2
 
From Taxonomies and Schemas to Knowledge Graphs: Part 3
From Taxonomies and Schemas to Knowledge Graphs: Part 3From Taxonomies and Schemas to Knowledge Graphs: Part 3
From Taxonomies and Schemas to Knowledge Graphs: Part 3
 
In Search of the Universal Data Model
In Search of the Universal Data ModelIn Search of the Universal Data Model
In Search of the Universal Data Model
 
Graph in Apache Cassandra. The World’s Most Scalable Graph Database
Graph in Apache Cassandra. The World’s Most Scalable Graph DatabaseGraph in Apache Cassandra. The World’s Most Scalable Graph Database
Graph in Apache Cassandra. The World’s Most Scalable Graph Database
 
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...
Enterprise Data Governance: Leveraging Knowledge Graph & AI in support of a d...
 
Graph Realities
Graph RealitiesGraph Realities
Graph Realities
 
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...
Powering Question-Driven Problem Solving to Improve the Chances of Finding Ne...
 
Semantic similarity for faster Knowledge Graph delivery at scale
Semantic similarity for faster Knowledge Graph delivery at scaleSemantic similarity for faster Knowledge Graph delivery at scale
Semantic similarity for faster Knowledge Graph delivery at scale
 
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...
Knowledge Graphs and AI to Hyper-Personalise the Fashion Retail Experience at...
 
Schema, Google & The Future of the Web
Schema, Google & The Future of the WebSchema, Google & The Future of the Web
Schema, Google & The Future of the Web
 
RAPIDS cuGraph – Accelerating all your Graph needs
RAPIDS cuGraph – Accelerating all your Graph needsRAPIDS cuGraph – Accelerating all your Graph needs
RAPIDS cuGraph – Accelerating all your Graph needs
 
Elegant and Scalable Code Querying with Code Property Graphs
Elegant and Scalable Code Querying with Code Property GraphsElegant and Scalable Code Querying with Code Property Graphs
Elegant and Scalable Code Querying with Code Property Graphs
 
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...
From Knowledge Graphs to AI-powered SEO: Using taxonomies, schemas and knowle...
 
Graph for Good: Empowering your NGO
Graph for Good: Empowering your NGOGraph for Good: Empowering your NGO
Graph for Good: Empowering your NGO
 
What are we Talking About, When we Talk About Ontology?
What are we Talking About, When we Talk About Ontology?What are we Talking About, When we Talk About Ontology?
What are we Talking About, When we Talk About Ontology?
 

Recently uploaded

Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptxLBM Solutions
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticscarlostorres15106
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machinePadma Pradeep
 
Bluetooth Controlled Car with Arduino.pdf
Bluetooth Controlled Car with Arduino.pdfBluetooth Controlled Car with Arduino.pdf
Bluetooth Controlled Car with Arduino.pdfngoud9212
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsHyundai Motor Group
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Enterprise Knowledge
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsMark Billinghurst
 
Pigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationSafe Software
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationSlibray Presentation
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksSoftradix Technologies
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsMemoori
 
costume and set research powerpoint presentation
costume and set research powerpoint presentationcostume and set research powerpoint presentation
costume and set research powerpoint presentationphoebematthew05
 
CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):comworks
 
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)Wonjun Hwang
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsPrecisely
 
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubUnleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubKalema Edgar
 

Recently uploaded (20)

Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptx
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 
Hot Sexy call girls in Panjabi Bagh 🔝 9953056974 🔝 Delhi escort Service
Hot Sexy call girls in Panjabi Bagh 🔝 9953056974 🔝 Delhi escort ServiceHot Sexy call girls in Panjabi Bagh 🔝 9953056974 🔝 Delhi escort Service
Hot Sexy call girls in Panjabi Bagh 🔝 9953056974 🔝 Delhi escort Service
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machine
 
Bluetooth Controlled Car with Arduino.pdf
Bluetooth Controlled Car with Arduino.pdfBluetooth Controlled Car with Arduino.pdf
Bluetooth Controlled Car with Arduino.pdf
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR Systems
 
Pigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food Manufacturing
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other Frameworks
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
 
costume and set research powerpoint presentation
costume and set research powerpoint presentationcostume and set research powerpoint presentation
costume and set research powerpoint presentation
 
CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):
 
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power Systems
 
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding ClubUnleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
 

Systems that learn and reason | Frank Van Harmelen

  • 1. Systems that learn and reason Frank van Harmelen Knowledge Representation & Reasoning Group Vrije Universiteit Amsterdam Creative Commons License CC BY 3.0: Allowed to copy, redistribute remix & transform But must attribute 1
  • 2. Direction of travel in modern AI “making artificial intelligence more human- centred” AI systems should “support people and be competent partners” But: “current AI systems are often incompetent because they lack background and contextual knowledge” 2
  • 3. So what’s holding AI back? For a long time, AI researchers have locked themselves in one of two towers 3 “current AI systems are often incompetent because they lack background and contextual knowledge” and “they cannot explain their actions”
  • 4. AI: a tale of two towers? Statistical AI Symbolic AI
  • 6. But our brain uses both … 6
  • 7. In our brain (roughly) Data Driven Knowledge Driven
  • 10. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff 10
  • 11. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff Very large knowledge graphs: Billions of facts and rules, But expensive to build & maintain: - Crowd sourcing 11
  • 12. Strengths & Weaknesses 10M training samples Symbolic Connectionist Construction Human effort Data hunger Scalable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff 4.8M training games 12
  • 13. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff worse with more data worse with less data 13
  • 14. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff
  • 15. Strengths & Weaknesses Symbolic Statistical Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff 15 “black box problem”
  • 16. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff quality  generality  16
  • 17. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff 17 Class: 793 Label: n04209133 (shower cap) Certainty: 99.7% Problem: “out of sample generalizability”
  • 18. Strengths & Weaknesses Symbolic Connectionist Construction Human effort Data hunger Scaleable +/- +/- Explainable + - Generalisable Performance cliff Performance cliff 18
  • 19. Can we get them to collaborate? 19
  • 20. Learning to add facts to knowledge graph (“knowledge graph completion”) Rolling Stones Angi Beat It Jackson Publish 20 Angi Rolling Stones Publish From: Predict: Yesterday Publish Beatles ….. Publish …..
  • 21. prediction algorithm 21 Learning to add facts to a knowledge graph (“knowledge graph completion”)
  • 22. Example: Inductive Logic Programming, rule mining (anyBURL on knowledge graphs) 22 parent(Mary,Vicky). parent(Mary,Andre). parent(Carrey,Vickey). mother(Mary,Vicky). mother(Mary,Andy). father(Carrey,Vickey). father(Carrey,Andy). parent(X,Y) :- mother(X,Y). parent(X,Y) :- father(X,Y) parent(Carrey,Andy).
  • 24. Informed ML using KR queen crown wears 24 shower cap ? Predict Select crown? showercap?
  • 25. Explainable ML using KR queen wears 25 shower cap ? Predict Explain crown?
  • 26.
  • 27. Context aware ML using KR (informed ML) See survey of 100+ systems in Von Rueden et al, Learning, 2019 flower? cushion? “Parts of a chair are: cushion and armrest” “Given the context of chair, a cushion is much more likely than a flower” P(cushion|chair) >> P(flower|chair)
  • 28. Learning intermediate abstractions :- see( , 3), see( , 5), add(3,5,8). End-to-end: 2x 784 inputs 19 outputs 8
  • 30. Learning intermediate abstractions Faster adaptation to changes; Better transfer learning
  • 31. Learning intermediate abstractions Faster adaptation to changes; Better transfer learning
  • 32. Injecting knowledge into rule-learning 32
  • 34. Injecting knowledge into rule-learning • Learn the rules 𝑟𝑑and linear terms l() • Learn the weights 𝛼 and 𝛽
  • 35. Injecting knowledge into rule-learning • Learn the rules 𝑟𝑑and linear terms l() • Experts give the rules 𝑟𝑐 and 𝑟𝑐 • Learn the weights 𝛼 and 𝛽
  • 37. 1. Machine Learning can benefit from injecting symbolic knowledge (knowledge graphs) • Explanations • Ranking hypotheses • Transfer learning • Sample efficiency • Pipeline engineering 2. The required symbolic knowledge is available at very large scale + corresponding tools “Why learn what you already know?” 37