MATHROCKS

Activities

CONGRESSES & WORKSHOPS

Coding Deep Neural Networks for PDEs

Workshop on HPC, Deep Learning, and Numerics in Geophysics

Zoom link: https://us06web.zoom.us/j/84540099896?pwd=MFFXTUJOVHA4YUZqN2RneFYwdzlzQT09

Workshop on Simulation and Inversion Methods in Geophysics

Workshop on Optimization Methods and Applied Mathematics

Location: National University of Colombia, Bogotá, Colombia.

Dates: Oct. 4-5, 2019.

Link: https://sites.google.com/view/womap/home

Congress on Mathematical and Computational Modeling of the Earth

Location: UPC, Barcelona, Spain.

Dates: May 23-24, 2019.

Oil & Gas Conference 2018, Bilbao Exhibition Center, Spain

M2OP (V) Fifth International Congress on Multiscale, and Optimization Problems

Location: BCAM, Bilbao, Spain.

Dates: May 24-25, 2018.

VMAD9

Location: Valparaíso, Chile.

Dates: January 17-18, 2019.

COURSES

Course

Coding Deep Neural Networks for PDEs

Course/Working group

Recent trends in deep learning research, and their possible applications on deep data-driven computing

Course/Working group

Solving Partial Differential Equations with Deep Learning

Solving Forward and Inverse Problems with Deep Learning

Deep Learning using TensorFlow 2.0

Mathematical Theory of Mixed Method for Linear Elasticity

Professor: Jay Gopalakrishanan (University of Portland, USA)

Place and time: BCAM, Bilbao, Spain (Sala Beta 2), June 4-8, 2018.

Adaptive Stabilized FEM

Professor: Victor M. Calo (Curtin University, Australia)

Place and time: BCAM, Bilbao, Spain (Sala Beta 2), May 18-22, 2020, 10:30-12:30 (10 hours course).

Description: Prof. Victor M. Calo will discuss several discretizations that build upon the ideas of finite elements and residual minimization to develop stable and adaptable numerical approximations. We use the mathematical framework developed by the discontinuous Galerkin community to build a new class of stabilized finite elements. Several model problems in 2D and 3D will serve to validate the presented theoretical results.

Programming with TF 2.0.

Professor: Artzai Picon (Tecnalia, Spain)

Place and time: BCAM, Bilbao, Spain (Sala Beta 2), Dec. 18th, 2019, 10:0-14:0 (4 hours course).

Description: Prof. Artzai Picon will discuss several aspects of TensorFlow 2.0 (TF2.0). He will show how to implement Deep Neural Networks (DNN) in TF2.0 for solving computational mechanics problems.

First Hands-on Experience with Deep Learning Algorithms

Professor: Dr. Adrián Galdrán (INESC TEC, Portugal)

Place and time: BCAM, Bilbao, Spain (Sala Beta_2) Wednesday, June 13th, 2018, 14:00-20:00 (6 hours).

Description: Dr. Galdrán will provide a short course with hands-on experience on how to install a Deep Learning framework and train deep neural networks on it. The course will verse about a specific library, Keras. Examples of how to use Keras for different problems will be reproduced and analyzed.

Introduction to Deep Learning for Computational Mechanics

Professor: Dr. Adrián Galdrán (INESC TEC, Portugal).

Location: BCAM, Bilbao, Spain. Dates: Jan 24-26, 2018.

Description: The aim of this course is to provide an introduction to deep learning algorithms for their use in computational mechanics. The instructor will introduce concepts like convolutional deep neural networks, supervised and unsupervised algorithms, activation functions to include nonlinearities (e.g, ReLU), data augmentation for teaching physical principles we know, the possibility of using existing libraries (e.g., those available in google, facebook), residual neural networks, and transfer learning.