February 2019 - Talks - lists.usi.ch

by Announcements of talks＠IDSIA

::When and where IDSIA's meeting room Monday 4th February h 15.45 ::Title Learning overcomplete and sparsifying transform with approximate and exact closed form solutions ::Abstract In this work, I addressed the learning problem for data-adaptive transform that provides sparse representation in a space with dimensions larger than (or equal to) the dimensions of the original space. I will present the iterative, alternating algorithm that has two steps: (i) transform update and (ii) sparse coding. In the transform update step, the focus is on a novel problem formulation based on a lower bound of the objective that addresses a trade-off between: (a) how much are aligned the gradients of the approximative objective and the original objective, and (b) how much the lower bound is close to the original objective. This allows us not only to propose approximate closed form solution, but it also gives the possibility to find an update that can lead to accelerated local convergence and enables us to estimate an update that can lead to a satisfactory solution under a small amount of data. Since in the transform update, the approximate closed form solution preserves the gradient and in the sparse coding step, we use exact closed form solution, the resulting algorithm is convergent. On the practical side, we evaluate on image denoising application and demonstrate promising denoising performance together with advantages in training data requirements, accelerations in the local convergence and the resulting computational complexity. ::Speaker Dimche Kostadinov PhD Student in Computer Science, University of Geneve

5 years, 1 month

1
0
0 / 0

h 10:00 "New statistical inference techniques for cosmology."

by Announcements of talks＠IDSIA

Guest:Julien Carron PostDoc (University of Sussex) Room G-222 (big room in Galleria 1) Time: starting at 10:15 Title: New statistical inference techniques for cosmology. Abstract: Cosmological science relies on observations of our Universe in order to try and answer some of fundamental physics most pressing questions. Within the next decade we will see large telescopes collecting data at unprecedented rates and resolution, where current analysis methods are provably un-sufficient. New analysis techniques including machine learning and superfast algorithms are mandatory for efficient inference on this new, large but structured data. In this talk I will discuss a couple of recent developments and open challenges on inverse problems from within the cosmology community, with direct applications to primordial gravitational waves measurements and the detection of the neutrino masses.

5 years, 1 month

1
0
0 / 0

talk on Monday, "New statistical inference techniques for cosmology."

by Announcements of talks＠IDSIA

Guest:Julien Carron, PhD Room G-222 (big room in Galleria 1) Time: starting at 10:15 Title: New statistical inference techniques for cosmology. Abstract: Cosmological science relies on observations of our Universe in order to try and answer some of fundamental physics most pressing questions. Within the next decade we will see large telescopes collecting data at unprecedented rates and resolution, where current analysis methods are provably un-sufficient. New analysis techniques including machine learning and superfast algorithms are mandatory for efficient inference on this new, large but structured data. In this talk I will discuss a couple of recent developments and open challenges on inverse problems from within the cosmology community, with direct applications to primordial gravitational waves measurements and the detection of the neutrino masses.

5 years, 2 months

1
0
0 / 0

PIDSIA this Friday by E. Merelli and M. Piangerelli

by Announcements of talks＠IDSIA

Title: Topology driven methods for complex systems modeling, an alternative approach to machine learning. Theory and some examples. Abstract: Many complex systems are characterized by multi-level properties that make the study of their dynamics and of their emerging phenomena a daunting task. The huge amount of data available in modern sciences can be expected to support great progress in these studies, even though the nature of the data varies. Given that, it is crucial to extract as much as possible features from data, including qualitative (topological) ones. The goal of TOPDRIM project has been to provide methods driven by the topology of data for describing the dynamics of multi-level complex systems. To this end, the project has developed new mathematical and computational formalisms accounting for topological effects. To pursue these objectives, the project brought together scientists from many diverse fields including as topology and geometry, statistical physics and information theory, computer science and biology. The proposed methods, obtained through concerted efforts, covered different aspects of the science of complexity ranging from foundations, to simulations through modelling and analysis, and constituted the building blocks for a new generalized theory of complexity. This seminar introduces the fundamentals behind the topological data analysis and through some applications developed both in the biomedical and financial field, presents the TOPDRIM methodology for going beyond the concept of networks by considering simplicial complexes instead. Speakers: Prof. Emanuela Merelli and Dr. Marco Piangerelli (University of Camerino, Computer Science Division) Host: Prof. Andrea Danani (IDSIA) Short bios of the speakers: Emanuela Merelli is a full professor of Computer Science at the University of Camerino. She founded the BioShape & Big Data laboratory conceived for young researchers with multidisciplinary interests in developing formal and computational methods to understand the complex behaviour of biological systems. She has been the coordinator of the TOPDRIM FP7-FET project and the partner of LITBIO (Laboratory for Interdisciplinary Technologies in Bioinformatics) Italian FIRB project. Currently, she is collaborating with Mario Rasetti for launching a new program on topological field theory of data with the aim to pave the way for automata synthesis as recognizers of data languages hidden in a given topological data space. Emanuela published many papers in refereed international journals and is continuously involved in the organization of events with an interdisciplinary character. Marco Piangerelli is a biomedical engineer, received his doctorate in Computer Science at the University of Camerino, under the supervision of Prof. Emanuela Merelli. He is currently a post-doc and contract professor at the same university where he deals with machine learning and topological data analysis. When: Friday, 15th February, 12:00-13:00 Location: Manno, Galleria 1, 2nd floor, room G1-201 Doodle registration: Pizza and drinks will be offered at the end of the talk. If you plan to attend, please register in a timely fashion at the following link so that we will have no shortage of food. https://doodle.com/poll/xx3zbunaehgm287u

5 years, 2 months

1
0
0 / 0

Tuesday 12 February. h 10:00: "Clustering non-stationary financial time series data."

by Announcements of talks＠IDSIA

IDSIA meeting room Tuesday 12 February. h 10:00: "Clustering non-stationary financial time series data." ::Abstract: Non-stationarity in data can arise due to the changes in various unobserved influencing factors. One way to account for non-stationarity is to employ models with time-varying parameters. Such models can be parametric or non-parametric depending on underlying assumptions they impose. The presented non-stationary approach identifies the optimal number of hidden regimes in data and the (a priori unknown) regime-switching dynamic without employing restrictive parametric assumption about the data-generating process. Within the regime, data is modelled using Maximum Entropy density, where the optimal number of density parameters is inferred via Lasso regularization technique. The resulting non-parametric methodology provides simultaneously the simplest and the least biased description of the data. ::Speaker Ganna Marchenko Institute of Computational Science - USI

5 years, 2 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

Talks February 2019