Legato Team aims at building intuitive and interactive platforms for computational mechanics problems which allow the users to interact with their models and hence gain insights into unconventional and counter-intuitive phenomena.
Very nice work... congratulations! The fact that you need the scale the strains for the PINN training is related to feature scaling. Feature scaling is very common on machine learning, especially when you have data with very different ranges. Quote "Feature standardization makes the values of each feature in the data have zero-mean (when subtracting the mean in the numerator) and unit-variance. This method is widely used for normalization in many machine learning algorithms (e.g., support vector machines, logistic regression, and artificial neural networks).[3][4]" (en.wikipedia.org/wiki/Feature_scaling).
Looks like block coordinate descent, where block corresponds to weights of the sub-problems defined by the speaker. Moreover, quasi newton can still be applied.
What does skew-symmetric mean in your case? It is a well-known term in statistics to refer to a specific type of asymmetric densities, but it is obviously not that here :)
What do you mean under "fast reanalysis"? Did you use information from previous analysis and perform iterative solution of subsequent analysis which converges quickly? Because at first glance these two analyses was independent
Great presentation!! I've been following you with so much passion and enthusiasm since 2018, I hope someday I will get the chance to join your amazing research team.
As usual, Vincent is extremely clear and explains to the audience what it is like to be a neurosurgeon and how data and computational sciences can impact such an important topic.