Zugzwang Meetings
2024-03-15 - IJCAR24 Reviews
Summary
- State-of-the-art:
- Thorough comparison with related work
- Motivation:
- Clarify the application of the approach
- Explore the advantages and limitations of the formalism
- Technical details:
- Self-containment
- Detail syntax and semantics of the considered class of programs.
- Clarify the relation of stable models and events
- Recall the stable model semantics and its properties
- Argument for Proposition 1 [is not] convincing
- Fixes:
- Provide the probabilities of the classes and of the events
- Clarify the role of "testing of the prior distributions"
- Give a general argument [about Bayesian networks] instead of an illustration on a simple example.
See Reviews file.
- Para ICLP24
- Mais técnico.
- Considerar scasp.
- Para KR24
- Mais formal.
2024-01-30 - Exploratory Research Project
Apply for FCT funding.
2024-01-05 - Next Research Lines
After the base-setting work of "An Algebraic Approach to Stochastic ASP" these are the next tasks to consider. Is summary:
- Logic Programming - Stratified & Non-stratified programs
- Computer Science - Inductive Logic Programming
- Software - Integration with Potassco and other frameworks
- Applications
Line 1: Logic Programming - Stratified & Non-stratified programs
Line 1a
Stratified & non-stratified programs are quoted in the "CREDAL" papers as important classes of logic programs.
Minimal example of a non-stratified program.
The following annotated LP, with clauses $c_1, c_2, c_3$ respectively, is non-stratified (because has a cycle with negated arcs) but no head is disjunctive:
0.3::a. % c1
b :- not c, not a. % c2
c :- not b. % c3
This program has three stable models: $$ \begin{aligned} m_1 &= \set{ a, c } \cr m_2 &= \set{ \neg a, b } \cr m_3 &= \set{ \neg a, c } \end{aligned} $$
We should investigate What are stratified programs and why are they important? and how does our approach deals with such programs?
Line 1b - Investigate the expressiveness of PASP
Consider:
- Recursion
- Variables,
- functional symbols,
Line 1c - The equivalence relation
Consider the cases where only $s \subseteq e$ and $e \subseteq s$. Or other refinements. Also consider the inconsistent and independent events.
Line 1d - Stability of the error function
Consider alternative error functions. See statistics, Kullback-Leibler divergence
Line 2: Computer Science - Inductive Logic Programming
Proceed from scoring programs to support genetic algorithms or other program space exploration methods.
Scoring programs, as described in our paper, is just a step into Inductive Logic Programming. To go further, we need to explore algorithms that:
- Use background knowledge, expressed as a PLP.
- Consult positive examples that should be soft induced.
- Consult negative examples that should be soft excluded.
- Generate PLPs that are scored.
- Recombine the best scored into a new population, using recombination rules.
In order to do that, PLPs must be expressed as data structures to be manipulated. Also recombination rules must investigated before become formally expressed and supported with adequate methods.
Line 3: Software - Integration with Potassco and other frameworks
Support annotated programs with zugzwang semantics.
- Bayesian Networks (BII Alice)
- Generate an annotated asp program from a bayesian network and run it trough
clingo
. - Recover the stable models from the previous ste and compute the respective probabilities.
- Generate an annotated asp program from a bayesian network and run it trough
- Program Manipulation
- Annotated ASP program representation and a parser.
Line 4: Applications
Apply zugzwang to a few showcases, besides the theoretic corner stones (non-stratified, disjunctive, bayes networks), preferably based in real world scenarios, with complex structure and large datasets.
- (Stochastic) Plan Generation
- Yale-Shooting Problem
- (Stochastic) Situation Calculus
- Frame Problem
- Latent Facts - and core assumptions.
- Given a Bayesian Network (or a Markov Networks):
- Represent it. (done for BNs; MNs?)
- Solve the common probability tasks: join (done), marginals, conditionals, parameter learning, inferring unobserved variables, sample generation, etc.
- Given a solved ASP specification:
- What is the marginal probability of the atom
a
? (done) - What other probability queries are important to consider?
- What is the marginal probability of the atom
- Given an unsolved ASP specification:
- What is the probability (distribution?) of the probabilistic fact
a
? - What other questions are relevant? E.g. the distribution family of a fact?
- What is the probability (distribution?) of the probabilistic fact
- Given a solved ASP specification and a set of samples:
- How do the probabilities inferred from the specification match the ones from the empiric distribution? (done might see alternative approaches)
- Given two solved ASP specification and a set of samples:
- Which specification best describes the empiric distribution? (done)
2024-01-05 - Publish Paper "AASASP"
Target conferences to publish paper "AASASP"
Conference | Abstract Deadline | Conference Date | Location | OBS |
---|---|---|---|---|
IJCAR 2024 | 2024-01-29 | 2024-07-3:6 | Nancy, France | ~Picked~ Rejected |
ECAI'24 | 2024-04-19 | 2024-10-19:24 | Santiago de Compostela, Spain | |
KR 2024 | 2024-04-24 | 2024-11-2:8 | Hanoi, Vietnam | |
GECCO 24 | 2024-02-05 | 2024-07-14:18 | Melboune, Australia | Overdue |
ICLP 24 | 2024-04-15 | preferred | ||
JELIA 25 | ||||
ICFP 24 | 2024-03-01 | 2024-09-2:7 | Milan, Italy |
2023-02-28 - Looking for Application Examples
What applications are we looking for?
- (Stochastic) Plan Generation
- Yale-Shooting Problem
- (Stochastic) Situation Calculus
- Frame Problem
- Given a Bayesian Network (or a Markov Networks):
- Represent it.
- Solve the common probability tasks: marginals, conditionals, parameter learning, inferring unobserved variables, sample generation, etc.
- Given a solved ASP specification:
- What is the marginal probability of the atom
a
? - What other probability queries are important to consider?
- What is the marginal probability of the atom
- Given an unsolved ASP specification:
- What is the probability (distribution?) of the probabilistic fact
a
? - What other questions are relevant? E.g. the distribution family of a fact?
- What is the probability (distribution?) of the probabilistic fact
- Given a solved ASP specification and a set of samples:
- How do the probabilities inferred from the specification match the ones from the empiric distribution?
- Given two solved ASP specification and a set of samples:
- Which specification best describes the empiric distribution?
What should be the task for the scholarship student? Use the Python
API of clingo
.
- Read a string and extract probability annotations; Associate those annotations with the respective atoms.
- Call
clingo
to get stable models. - Support computation of the equivalence classes: Which functions and relations?
- Compute event probability using weighted model counting on the equivalence classes.
- Read a Bayesian Network from a file (
BIF
,DSC
,NET
,RDA
,RDS
, ...) and generate an annotated "ASP" specification.
2022 - AAAI - Inference and Learning with Model Uncertainty in Probabilistic Logic Programs
- Is "Epistemic Uncertainty (EU)" the right framework for Zugzwang? How relevant are the epistemic questions in this paper to our work?
- EU can be represented by Credal Sets, Subjective Logic and Beta Distributions?
- Experiments made with BNs from (Kaplan and Ivanovska 2018) and larger networks from the BNLearn repository.
- Are networks, Bayesian Networks in particular, a "good enough" pool of "example applications" to us, for now?
2023-01-10 - 15:00
- Paper
- Project
- Latent Facts
2022-12-12
- Is the project proposal ok? How long/detailed should it be?
- Initial exploratory code
event_lattice.py
andEventLattice.ipynb
done. - Start writing paper: Introduction, state of the art, motivation
- Next task for prototype:
- Get stable models from potassco/s(casp)
- other?
2022-12-05
- Created shared folder (gdrive:zugzwang) https://drive.google.com/drive/folders/1xs-cjxWJzn2JxqeNgh9LX5xWN50BW-Be?usp=share_link
- Refine project tasks, for Bachelor, M.Sc., Ph.D. students and for researchers.