Teaching code

4S010683

Teacher

Daniele Meli

Coordinator

Daniele Meli

Credits

6

Language

English

Scientific Disciplinary Sector (SSD)

INF/01 - INFORMATICS

Period

Semester 1  dal Oct 1, 2024 al Jan 31, 2025.

Courses Single

Authorized

Lessons timetable Moodle Seminars 1

Learning objectives

This course builds on knowledge about statistical/machine/deep learning methods and aims at providing means to understand the “why” and the “how” of their outcomes. After introducing the basic concepts, a taxonomy of the existing methods will be provided, then the main state-of-the-art approaches for neurosymbolic AI will be illustrated. The theoretical part will be complemented by practical sessions where the concepts that have been acquired will be put in practice considering specific case-studies.

At the end of the course the students will have acquired fundamental skills about explainability, interpretability, randomness and causality; the knowledge of the main methods for interpretability (intrinsic methods, post-hoc, model-specific, model-agnostic, local, global, etc.), of the related properties (sensibility, implementation invariance, separability, stability, completeness, correctness, compactness), of the main types of explanations and their properties (accuracy, fidelity, consistency, stability, comprehensibility, certainty and relevance), and of the main visualization methods (activation maps, LRP, GradCam). Additionally, students will need to demonstrate knowledge of state-of-the-art approaches to neuro-symbolic artificial intelligence, with main focus on: standard deep learning; symbolic solvers that use neural networks as sub-routines for state estimation; hybrid systems with neural network and symbolic system specialized on complementary tasks with interaction through input/output; symbolic knowledge compiled in the training set of a neural network; neural computing systems that contain symbolic reasoning systems (type 1 and 2 reasoning).

Examination methods

To pass the exam, students must demonstrate:
- to have understood the theoretical and methodological aspects of the teaching
- to know how to apply the acquired knowledge to solve application problems presented in the form of exercises, questions and projects.

Prerequisites and basic notions

Model-free and model-based RL; basics of statistics; basics of automated reasoning and propositional / first-order / temporal logics.

Program

explainability and interpretability for AI; concepts and algorithms for the causal analysis of data in the form of time series (Granger causality, main assumptions of causality, main algorithms including PCMCI), with application to the identification of patterns and the identification of anomalies; autonomous planning based on formal methods (logic programming, answer set programming); learning of logical explanations from data (inductive logic programming, induction in the semantics of answer sets), with application to the interpretation of policies for reinforcement learning agents; neurosymbolic planning and learning, combining reinforcement learning techniques with techniques based on logic programming and logic induction; principles of explainable human-machine interaction (human-machine collaboration, extraction of explainable patterns from the interaction with the machine).

Didactic methods

Almost all theoretical lectures will be linked to lab sessions with the computer for practical implementation of concepts and algorithms, with the support of a teaching assistant

Learning assessment procedures

The exam will consist of 2 parts;
1. theoretical interview XOR practical project or study of a research paper (in agreement with the teacher)
2. presentation of lab assignments

Evaluation criteria

theoretical and implementative skills acquired regarding the topics of lab and theoretical lessons.

Criteria for the composition of the final grade

60% theory / project / paper; 40% lab assignments

Exam language

inglese / english