EAD model
Topics to be chosen for presentation plus instructions
Classes
April 2nd
Knowledge Representation (PT) (EN) (Recorded Class (PT))
Exercises (Solutions)
Summary: Knowledge
representation. Data x Knowledge. Various ways of representing
data and knowledge: natural language, databases, programming
languages and data structures, scripts, frames, semantic
networks, knowledge graphs, logic
For next class: Read:
Relational inductive
biases, deep learning, and graph networks
April 9th: Easter Holiday
April 16th
Worksheet #1
(suggested answers)
Probabilistic Logic Programming (Recorded Class)
Practical
Summary: ~1h to answer worksheet
#1. Discussion. Limitations of graph
representations. Probabilistic Logic
Programming. ProbLog. Worlds. Inference. Basic examples of ProbLog.
For next class: Read:
Inference and learning in probabilistic logic programs using weighted boolean formulas
April 23rd
Worksheet #2 (30 min for reading and head start, send me answers by email, please) (suggested answers)
Relational Machine Learning (up to slide 13 + 25-35) (aleph run) (Recorded Class)
Summary: ~30 minutes to have a look at the worksheet and head start. Discussion about questions posed by students by email (mostly about section 3 of GraphNets paper). Introduction to Inductive Logic Programming (ILP).
For next class: Read:
Inductive Logic Programming
ILP turns 20
Turning 30: New Ideas in Inductive Logic Programming
April 30th
Worksheet #3 (30 min for reading and head start, send me answers by email, please) (suggested answers)
Relational Machine Learning (cont.) (Recorded Class)
Practical: learning first order rules with Aleph
Summary: ~30 minutes to have a look at the worksheet and head start. Relational Learning using Aleph. Practical: learning parent and grandparent.
For next class: Read:
Neural Networks for Relational Data
May 7th
Worksheet #4 (30 min for reading and head start, send me answers by email, please) (suggested answers)
Relational Machine Learning (cont.) (Recorded Class)
Markov Models
Markov Networks
Summary: ~30 minutes to have a look at the worksheet and head start. Learning rules: using refinement operators and controlling the search space. Markov chain and Markov networks.
For next class: Read:
Scalable Neural Methods for reasoning with a symbolic knowledge base
May 14th
Worksheet #5 (cancelled)(30 min for reading and head start, send me answers by email, please)
Markov Networks (cont.) (Recorded Class)
Markov Logic Networks
Practical: Playing with Markov Logic Networks using alchemy-2
Summary: ~30 minutes to have a look at the worksheet and head start. Markov logic networks.
May 21st
May 28th