15 April 2024 - 18 April 2024

Dr. Federico Fusco, from Sapienza University of Rome, is going to teach a Ph.D. level class on "Submodular Optimization", 10 hours, 15th-18th April. The class is open to all the interested Ph.D. students (but more senior colleagues are very welcome to join). At the end there will be an optional graded homework for the interested Ph.D. students (which might grant 1ECTS if passed).
East Campus USI-SUPSI

15 March 2024 - 15 March 2024

The topic of this second seminar will be “Causal Inference (not only) in Clinical Trials”. The topic will be introduced by three short talks (~10 min each) by Radka Švihrová, Michal Bechny, and Alvise De Rossi, PhD students @MeDiTech. After these three short talks there will be an open discussion on the topic.
Room B1.07

24 January 2024 - 24 January 2024

The Premurosa project, funded by the EU, has made significant use of machine learning to examine data and offer individualized insights on various aspects of medicine and biology. This includes the use of spiking neural networks to predict toxicity, bioaccumulation, and enzyme activity, as well as data fusion to interpret biomarkers or characterize biomaterials, and anomaly detection in proteomics.
Room C2.09, Sector C, East Campus USI-SUPSI

12 January 2024 - 12 January 2024

SARS-CoV-2 was declared a pandemic by the WHO on March 11th, 2020. Public protective measures were enforced in every country to limit the diffusion of SARS-CoV-2. Its transmission, mainly by droplets, has been measured by the effective reproduction number (Rt) that counts the number of secondary cases caused in a population by an average infectious individual at time t. Current strategies to calculate Rt reflect the number of secondary cases after several days, due to a delay from symptoms onset to reporting. We propose a complementary Rt estimation using supervised machine learning techniques to predict short term variations with more timely results.
IDSIA Large meeting room

21 December 2023 - 21 December 2023

This mini-workshop on the philosophy of machine learning sees the participation of a group of selected scholars who will discuss topics such as the role of justification in machine learning algorithms and a novel role for explainable artificial intelligence as a "mediator". More information and details in the programme here below.
Room C2.09, Sector C, East Campus USI-SUPSI

28 November 2023 - 28 November 2023

Reassembly tasks are fundamental human skills acquired during early developmental stages; therefore, we believe it is necessary to develop these tasks to approach general artificial intelligence (AI). The resolution of reassembly tasks in both 2D and 3D holds significant relevance across various fields such as biology, computer vision, and cultural heritage. However, existing approaches tend to focus solely on specific tasks and modalities, lacking a unified framework. In this presentation, we will introduce a novel unified framework based on a Graph Neural Network architecture. We will also explore the benefits and issues related to the adoption of a diffusion process, where we introduce noise into the elements' positions and orientations, followed by iterative denoising to reconstruct their coherent poses. Through our study, we will reveal the shared fundamentals between 2D and 3D tasks, emphasizing the significance of rotation-equivariant representation as a common inductive bias that enhances performance in both modalities.
East Campus USI-SUPSI, Room B1.09

23 November 2023

Causal inference from observational data is a compelling problem in statistics, which has attracted much attention due to its potential application in various scientific fields. Estimating the effects of a manipulation on a system of random variables however poses both modeling and computational challenges, which are typically addressed by imposing strict assumptions on the joint distribution such as linearity. One appealing approach is to model the system as a Gaussian process network (GPN), which allows describing the causal relationships among a set of random variables with minimal parametric assumptions. In the absence of prior knowledge of the underlying causal graph, a fully Bayesian approach requires integrating the causal quantity of interest over the posterior over graphs, which is computationally infeasible even in low dimensions. By harnessing Monte Carlo and Markov Chain Monte Carlo methods we can sample from the posterior distribution of network structures, thus providing an accurate approximation of the posterior. Causal inference across the whole GPN can then be performed while also accounting for uncertainty in the causal graph. Simulation studies show that our approach is able to identify the effects of hypothetical interventions with non-Gaussian, non-linear observational data and accurately reflect the posterior uncertainty of the causal estimates. Finally we compare the results of our GPN-based causal inference approach to existing methods on a real dataset of A. Thaliana gene expressions.
East Campus USI-SUPSI, Room C2.09

21 November 2023 - 21 November 2023

The Euclidean Steiner tree problem requires a shortest network interconnecting a given set of points in the plane. Additional vertices may be introduced and are called Steiner points. This is a well-studied, but NP-hard problem. Nevertheless, the current flagship algorithm can exactly solve instances on many thousands of input points. There are many variations to this classical problem. In this presentation we will look at a multi-source multi-sink directed version. For two given sets of points A (the sources) and B (the sinks), the task is to find a minimum length network such that there exists a directed path between every source and every sink. We will share some known structural results on optimal solutions and discuss the current state of algorithmic approaches for finding exact solutions.
East Campus USI-SUPSI Room B1.14

10 November 2023 - 10 November 2023

For a language model (LM) to faithfully model human language, it must compress vast, potentially infinite information into a relatively low-dimensional space. On this topic, I will present a recent work with Corentin Kervadec and Marco Baroni to appear at EMNLP. We propose analyzing compression in (pre-trained) LMs from two points of view: geometric and information- theoretic. We demonstrate that the two views are highly correlated, such that the intrinsic geometric dimension of linguistic data predicts their coding length under the LM. We then show that, in turn, high compression of a linguistic dataset predicts rapid adaptation to that dataset, confirming that being able to compress linguistic information is an important part of successful LM performance. As a practical byproduct of our analysis, we evaluate a battery of intrinsic dimension estimators for the first time on linguistic data, showing that only some encapsulate the relationship between information-theoretic compression, geometric compression, and ease-of-adaptation.
Room D1.14 - East Campus USI-SUPSI

2 November 2023 - 3 November 2023

The aim of the meeting, organised under the auspices of IDSIA USI-SUPSI, the College of Humanities at EPFL and the Digital Society Initiative at the University of Zurich, is to bring together experts from different fields such as philosophy, bioethics, AI ethics, XAI, and human-computer interaction to discuss if, how and to what extent proposed solution to the black-box problem are effective in supporting the successful integration and appropriation of AI systems in medical practice.
East Campus USI-SUPSI - Room C1.02

The notion of trust is a major player, in many epistemic and computational contexts, among others. Such notion appears especially relevant in all those situations where verification or evaluation for knowledge is missing, not reachable or non-existent, and agents must rely on information received by others. This includes cases where expert knowers may not yet ground their claims, and the public has to build an opinion on the subject matter by considering the dynamic of the information exchange. Formal logic approaches to this aim have been increasingly important and diverse in the last decades. Recently, the notion of negative trust has been formalized in terms of a proof theory and semantics: this approach distinguishes between distrust as the epistemic act of rejecting incoming contradictory information considered unreliable; and mistrust as the epistemic act of updating one's own information state by removing previously held data in order to accommodate newly received information, considered more reliable or up to date. A natural extension of such multi-agent contexts is where epistemic acts are performed under uncertainty, at several levels: the claims of the agents may be graded; information may reach agents with a certain degree of probability; and the degree of acceptance or rejection of the information received may not be binary. In this talk we present a logic of negative trust applied to uncertain judgements. We offer a proof theory and a relational semantics for which standard soundness and completeness results hold. This is a joint work with Francesco Antonio Genco and Giuseppe Primiero.
Room D1.01 (Campus-Est),