4 papers
PolyNODE: Variable-dimension Neural ODEs on M-polyfolds
Per à hag, Alexander Friedrich, Fredrik Ohlsson +1
Neural ordinary differential equations (NODEs) are geometric deep learning models based on dynamical systems and flows generated by vector fields on manifolds. Despite numerous suc…
Explicit Solutions in Isotropic Planar Elastostatics
Andreas Granath, Per à hag, Antti Perälä +1
Addressing the intricate challenges in plane elasticity, especially with non-vanishing traction and complex geometries, requires innovative methods. This paper offers a novel appro…
Kiselman Minimum Principle and Rooftop Envelopes in Complex Hessian Equations
Per à hag, RafaŠCzyż, Chinh H. Lu +1
We initiate the study of -subharmonic functions with respect to a semipositive -form in Euclidean domains, providing a significant element in understanding geodesics with…
Geodesic connectivity and rooftop envelopes in the Cegrell classes
Per à hag, RafaŠCzyż, Chinh H. Lu +1
This study examines geodesics and plurisubharmonic envelopes within the Cegrell classes on bounded hyperconvex domains in . We establish that solutions possessing com…