Learning any memory-less discrete semantics for dynamical systems represented by logic programs

Ribeiro, Tony; Folschette, Maxime; Magnin, Morgan; Inoue, Katsumi

doi:10.1007/s10994-021-06105-4

Cited by 9 publications

(22 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [31,33], LFIT was extended to learn systems dynamics independently of its update semantics. That extension relies on a modeling of discrete memory-less multi-valued systems as logic programs in which each rule represents that a variable possibly takes some value at the next state, extending the formalism introduced in [12,34].…”

Section: Learning From Interpretation Transition (Lfit)mentioning

confidence: 99%

“…That extension relies on a modeling of discrete memory-less multi-valued systems as logic programs in which each rule represents that a variable possibly takes some value at the next state, extending the formalism introduced in [12,34]. The representation in [31,33] is based on annotated logics [4,3]. Here, each variable corresponds to a domain of discrete values.…”

Section: Learning From Interpretation Transition (Lfit)mentioning

confidence: 99%

“…GULA [31,33] and PRIDE [32] are particular implementations of the LFIT framework [12]. In the present section we introduce notation and describe the fundamentals of both methods.…”

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

“…In [33], the set of variables is divided into two disjoint subsets: T (for targets) and F (for features). It allows to define dynamic MVLP which capture the dynamics of the problem we tackle in this paper.…”

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

“…GULA [31,33] and PRIDE [32] can produce such programs. Formally, given a set of observations T , GULA [31,33] and PRIDE [32] will learn a set of rules P such that all observations are explained: ∀(s, s ) ∈ T, ∀v val ∈ s , ∃R ∈ P, R s, h(R) = v val . All rules of P are correct w.r.t.…”

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

See 4 more Smart Citations

Symbolic AI for XAI: Evaluating LFIT Inductive Programming for Fair and Explainable Automatic Recruitment

Ortega¹,

Fiérrez²,

Morales³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning methods are growing in relevance for biometrics and personal information processing in domains such as forensics, e-health, recruitment, and e-learning. In these domains, white-box (human-readable) explanations of systems built on machine learning methods can become crucial. Inductive Logic Programming (ILP) is a subfield of symbolic AI aimed to automatically learn declarative theories about the process of data. Learning from Interpretation Transition (LFIT) is an ILP technique that can learn a propositional logic theory equivalent to a given blackbox system (under certain conditions). The present work takes a first step to a general methodology to incorporate accurate declarative explanations to classic machine learning by checking the viability of LFIT in a specific AI application scenario: fair recruitment based on an automatic tool generated with machine learning methods for ranking Curricula Vitae that incorporates soft biometric information (gender and ethnicity). We show the expressiveness of LFIT for this specific problem and propose a scheme that can be applicable to other domains.

show abstract

Section: Learning From Interpretation Transition (Lfit)mentioning

confidence: 99%

Section: Learning From Interpretation Transition (Lfit)mentioning

confidence: 99%

“…GULA [31,33] and PRIDE [32] are particular implementations of the LFIT framework [12]. In the present section we introduce notation and describe the fundamentals of both methods.…”

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

Section: Pride Implementation Of Lfitmentioning

confidence: 99%

See 3 more Smart Citations

Symbolic AI for XAI: Evaluating LFIT Inductive Programming for Fair and Explainable Automatic Recruitment

Ortega¹,

Fiérrez²,

Morales³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Condition for Periodic Attractor in 4-Dimensional Repressilators

Sun,

Folschette,

Magnin

2023

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

One of the key questions about gene regulatory networks is how to predict complex dynamical properties based on the influence graph's topology. Earlier theoretical studies have identified conditions for complex dynamical properties, like multistability or oscillations, based on topological features, like the presence of a positive (negative) feedback loop. This work follows this path and aims to find a sufficient and necessary condition for the existence of a periodic attractor in 4-dimensional (4 genes) repressilators based on a discrete modeling framework under some dynamical assumptions. These networks are extensions of the widely studied 3-dimensional repressilator, which has been used in synthetic biology to produce synthetic oscillations. While other researchers have explored specific extensions of the 3-dimensional repressilator to improve synthetic oscillation control, our work investigates all 4-dimensional networks with only inhibitions. By uncovering new insights about periodic attractors in these small networks, our findings could aid the design of new synthetic oscillations. We search for condition for period attractor in an exhaustive manner with the guide of a decision tree model. Our major contributions include: 1) discovering that, with one exception, the relations between gene regulation thresholds do not impact the existence of periodic attractors in any of the influence graphs considered in this study; 2) identifying a sufficient and necessary condition of simple form for the existence of a periodic attractor when the exception is ignored; 3) identifying new topological features of influence graphs that are necessary for predicting the existence of periodic attractor in 4-dimensional repressilators.

show abstract

Diagnosis of Event Sequences with LFIT

Ribeiro,

Folschette,

Magnin

et al. 2024

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Diagnosis of the traces of executions of discrete event systems is of interest to understand dynamical behaviors of a wide range of real world problems like real-time systems or biological networks. In this paper, we propose to address this challenge by extending Learning From Interpretation Transition (LFIT), an Inductive Logic Programming framework that automatically constructs a model of the dynamics of a system from the observation of its state transitions. As a way to tackle diagnosis, we extend the theory of LFIT to model event sequences and their temporal properties. It allows to learn logic rules that exploit those properties to explain sequences of interest. We show how it can be done in practice through a case study.

show abstract

Learning any memory-less discrete semantics for dynamical systems represented by logic programs

Cited by 9 publications

References 49 publications

Symbolic AI for XAI: Evaluating LFIT Inductive Programming for Fair and Explainable Automatic Recruitment

Symbolic AI for XAI: Evaluating LFIT Inductive Programming for Fair and Explainable Automatic Recruitment

Condition for Periodic Attractor in 4-Dimensional Repressilators

Diagnosis of Event Sequences with LFIT

Contact Info

Product

Resources

About