Adaptive Machine Learning for Data Streams

25 enero, 2021
9:00 am

Título:  Adaptive Machine Learning for Data Streams

Ponente: Albert Bifet (University of Waikato)

Organizador: Alejandro Rabasa

Fecha: Lunes 25 de enero de 2021 a las 9:00 horas.

PINCHA AQUÍ PARA ACCEDER

Abstract: Big Data and the Internet of Things (IoT) have the potential to fundamentally shift the way we interact with our surroundings. The challenge of deriving insights from the Internet of Things (IoT) has been recognized as one of the most exciting and key opportunities for both academia and industry. Advanced analysis of big data streams from sensors and devices is bound to become a key area of data mining research as the number of applications requiring such processing increases. Dealing with the evolution over time of such data streams, i.e., with concepts that drift or change completely, is one of the core issues in stream mining. In this talk, I will present an overview of data stream mining, and I will introduce some popular open source tools for data stream mining.

Seminario Online Senén Barro Ameneiro

Título:    Inteligencia Artificial en una sociedad de dispositivos

Ponente: Senén Barro Ameneiro (Universidad de Santiago de Compostela)

Organizador: Federico Botella

Fecha: Lunes 21 de diciembre de 2020 a las 12:00

PINCHA AQUÍ PARA VER EL SEMINARIO

ABSTRACT: Vivimos en una sociedad de dispositivos. Por una parte, los dispositivos incorporan una creciente capacidad de percepción, computación, interacción y acción y conforman sociedades de entidades cada vez más interconectadas. El caso de los teléfonos móviles es paradigmático. Por otra parte, nuestra sociedad, la humana, es cada vez más dependiente de ellos. La Inteligencia Artificial es a la vez una necesidad y una oportunidad para esta sociedad de dispositivos. Se hace necesaria para gestionar una ingente cantidad de datos e información y al tiempo se abre un sinfín de posibilidades derivadas del aprendizaje automático colectivo por parte de las máquinas. Pondremos ejemplos derivados del geoposicionamiento de móviles en interiores, la lucha contra la pandemia o la monitorización de nuestra salud.

Digital Humanities – Theorie und Methodik “Data Literacy and Digital Humanities”

Título:    «Data Science and Digital Humanities: a feasible merge?»

Ponente:  Alejandro Bia (Centro de Investigación Operativa)

Organizador: Prof. Elisabeth Burr de la Universidad de Leipzig

Fecha: Martes 8 de diciembre de 2020 a las 17:00

INSCRIPCIÓN INSCRIPCIÓN

ABSTRACT: This talk deals with the new and trendy field of Data Science, tracing back to its origins and analysing its future potential applications especially within the Digital Humanities. We will see the different subjects Data Science comprise and the different roles and backgrounds of the people who work in this field. We will also compare and discuss the relationships and possible collaborations, both in terms of research and education, between these two interdisciplinary bodies of knowledge.

Seminario Online Judit Muñoz Matute

Título:    The application of Discontinuous Petrov-Galerkin method for time-dependent Partial Differential Equations.

Ponente:  Judit Muñoz Matute (BCAM – Basque Center for Applied Mathematics)

Organizador: José Valero

Fecha: Lunes 14 de diciembre de 2020 a las 12:00

PINCHA AQUÍ PARA VER EL SEMINARIO

ABSTRACT: The Discontinuous Petrov-Galerkin (DPG) method with optimal test functions is a numerical method for approximating the solution of Partial Differential equations. It was proposed 10 years ago and since then, it has been applied to the simulation of a wide variety of problems including convection-dominated diffusion, Maxwell’s equations, linear elasticity, Stoke’s flow and Helmholtz equation, among many others. The key idea of the DPG method is to construct optimal test functions in such a way that the discrete stability is inherited from the continuous problem. In this talk, I will show how to apply the DPG method in the time variable for transient PDEs and its relation with exponential time integrators.

Seminario Online Dolores Romero Morales

7 diciembre, 2020
12:00 pm

Título:    Explaining Machine Learning Outcomes by Means of Mathematical Optimization

Ponente:  Dolores Romero Morales (Copenhagen Business School, Denmark)

Organizador: Juan Fco. Monge

Fecha: Lunes 7 de diciembre de 2020 a las 12:00

PINCHA AQUÍ PARA VER EL SEMINARIO

ABSTRACT: There is a growing literature on enhancing the interpretability of Machine Learning methods involved in Data Driven Decision Making. Interpretability is desirable for non-experts; it is required by regulators for models aiding, for instance, credit scoring; and since 2018 the European Union has extended this requirement by imposing the so-called right-to-explanation in algorithmic decision making. Mathematical Optimization has shown a crucial role when striking a balance between interpretability and accuracy, having LASSO as one of the main exponents. In this presentation, we will go through very recent examples of enhancing the interpretability of Supervised as well as Unsupervised Learning methods with the help of Mixed Integer NonLinear Optimization.

Seminario Online Francesco Ciardiello

30 noviembre, 2020
12:00 pm

Título:   On Pure-Strategy Nash Equilibria in a Duopolistic Market Share Model

Ponente: Francesco Ciardiello (Sheffield University Management School)

Organizador: Ana Meca

Fecha: Lunes 30 de noviembre de 2020 a las 12:00

PINCHA AQUÍ PARA VER EL SEMINARIO

ABSTRACT: This paper develops a duopolistic discounted marketing model with linear advertising costs and advertised prices for mature markets still in expansion. Generic and predatory advertising effects are combined together in the model. We characterize a class of adver- tising models with some lowered production costs. For such a class of models, advertising investments have a no-free-riding strict Nash equilibrium in pure strategies if discount rates are small. We discuss the entity of this efficiency at varying parameters of our advertising model. We provide a computational framework in which market shares can be computed at equilibrium, too. We analyze market share dynamics for an asymmetrical numerical scenario where one of the two firms is more effective in generic and predatory advertising. Several numerical insights on market share dynamics are obtained. Our computational framework allows for different scenarios in practical applications and it is developed, thanks to Mathematica software.