Seminari - Dipartimento Computer Science validi dal 19.04.2024 al 19.04.2025. https://www.di.univr.it/?ent=seminario&rss=0&lang=en https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6266 Relatore: Thiago D. Simão; Provenienza: Eindhoven University of Technology; Data inizio: 2024-05-29; Ora inizio: 12.30; Referente interno: Alberto Castellini; Riassunto: Safety is a crucial concern when deploying reinforcement learning (RL) algorithms in real-world scenarios. In this two-part lecture series, we delve into safety considerations from two perspectives: ensuring reasonable performance and adhering to predefined constraints. - PART 1. In the first segment, we investigate the offline setting where the RL agent solely accesses a fixed dataset of prior trajectories, devoid of direct interaction with the environment. Given the availability of the behavior policy responsible for data collection, the primary challenge is crafting a policy that outperforms such behavior policy. We study algorithms that leverage the behavior policy to compute an improved policy with high probability and discuss how to improve their sample efficiency. - PART 2. Transitioning to the latter segment, we confront the limitations inherent in specifying the behavior expected from an agent solely via a reward function. We introduce a model that mitigates this issue using constraints, and we discuss how to compute the corresponding optimal policy when the problem is known. Finally, we study algorithms that can efficiently explore the environment and eventually converge to an optimal policy when the model is unknown. Schedule: May 29, 12.30-14.30 (Room B) May 29, 15.30-17.30 (Room 1.02) May 31, 15.30-17.30 (Room 1.02) June 5, 12.30-14.30 (Room B) June 5, 15.30-17.30 (Room 1.02 June 7, 15.30-17.30 (Room 1.02) The minicourse is related to the quot;Reinforcement Learningquot; course (Master in Artificial Intelligence). Wed, 29 May 2024 12:30:00 +0200 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6266 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6279 Relatore: Riccardo Volpi; Provenienza: Naver Labs Europe - France; Data inizio: 2024-05-28; Ora inizio: 14.00; Note orario: Sala Verde; Referente interno: Vittorio Murino; Riassunto: In definizione. Tue, 28 May 2024 14:00:00 +0200 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6279 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6070 Relatore: Silvia Lavagnini; Provenienza: Oslo University - Department of Data Science and Analytics Nydalsveien; Data inizio: 2024-05-13; Referente interno: Luca Di Persio; Riassunto: In this mini-course we will discuss various numerical methods for the pricing of financial instruments, with a particular focus on applications to the energy markets. We will start from the simulations of some of the most common continuous-time models, such as the geometric Brownian motion and the Ornstein-Uhlenbeck process. Stochastic volatility models, such as the Heston model, may also be discussed. We will then consider different approaches for options pricing, such as the PDE and the Monte Carlo approach. State-dependent options may also be discussed. Finally, forward contracts with delivery period typical of the energy markets will be treated from the Heath-Jarrow-Morton modelling point of view. If time allows, we will also discuss (simulations of) SPDEs for Hilbert space-valued forward curves. The course will be given in Python. Mon, 13 May 2024 00:00:00 +0200 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6070 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6256 Relatore: Daniela Bubboloni; Provenienza: Università di Firenze; Data inizio: 2024-05-09; Ora inizio: 15.30; Note orario: Aula L (presenza e remoto); Referente interno: Lidia Angeleri; Riassunto: Da definire Link:_https://unipd.zoom.us/j/82518660070?pwd=RUpxL1FnZG9yVzFrOCtrM0xYMEZaZz09_ Meeting ID: 825 1866 0070 Password: 62542 Referente: Lidia Angeleri . Thu, 9 May 2024 15:30:00 +0200 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6256 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6276 Relatore: Franco Zivcovich; Provenienza: Neurodec, Nice, France; Data inizio: 2024-05-06; Ora inizio: 12.30; Note orario: starting date; Referente interno: Marco Caliari; Riassunto: In this course you will gain a deep understanding of Finite Element Methods (FEM) using Python. We will start by revisiting sparse data structures and iterative solvers, utilizing popular libraries like Numpy, Scipy, and high-performance packages such as PETSc. You will then delve into unstructured simplicial meshes, learn to work within the reference element, and employ meshes for approximating integrals over complex domains. Furthermore, you will develop expertise in piecewise approximation and master the assembly of matrices crucial in solving Partial Differential Equations (PDEs) with Finite Elements. By completing this course, you will not only grasp the mechanics of FEM at an unparalleled depth but also acquire the skills to tailor specific solutions for unique engineering problems. This knowledge will set you apart from engineers who rely solely on off-the-shelf, compiled packages like FEniCS and FreeFEM++, allowing you to achieve unmatched performance in your engineering endeavors. Join us in this course and embark on a journey to become a proficient engineer capable of navigating the complex world of mathematical modeling and computational engineering. Timetable: May 6, 12:30-14:30, room G May 8, 9:30-11:30, meeting room second floor May 9, 9:30-11:30, room L May 20, 12:30-14:30, room G May 22, 16:30-18:30, room L May 23, 9:30-11:30, room L . Mon, 6 May 2024 12:30:00 +0200 https://www.di.univr.it/?ent=seminario&rss=0&lang=en&id=6276