List your company

Function Approximation on F2

0MSEJacobi-rKAN
-2,212,000,00012,719,000,00027,650,000,00042,581,000,000Nov 7, 2019Aug 15, 2020May 25, 2021Mar 4, 2022Dec 12, 2022Sep 21, 2023Jun 30, 2024

Evaluation results

MethodLinks
Jacobi-rKANK (degree or grid size)=22024.060—
Padé[K/3]-rKANK (degree or grid size)=22024.060—
Tanh2024.060—
Padé[K/3]-rKANK (degree or grid size)=32024.060—
fKANK (degree or grid size)=32024.060—
fKANK (degree or grid size)=22024.060—
Padé[K/4]-rKANK (degree or grid size)=32024.060—
Padé[K/3]-rKANK (degree or grid size)=52024.060—
Padé[K/3]-rKANK (degree or grid size)=42024.060—
Jacobi-rKANK (degree or grid size)=32024.060—
ChebNetdegree n=442019.110—
Padé[K/6]-rKANK (degree or grid size)=52024.060—
Jacobi-rKANK (degree or grid size)=42024.060—
ChebNetdegree n=402019.110—
Padé[K/4]-rKANK (degree or grid size)=42024.060—
Padé[K/4]-rKANK (degree or grid size)=52024.060—
fKANK (degree or grid size)=42024.060—
Padé[K/3]-rKANK (degree or grid size)=62024.060—
ChebNetdegree n=362019.110—
fKANK (degree or grid size)=52024.060—
fKANK (degree or grid size)=62024.060—
ChebNetdegree n=322019.110—
Jacobi-rKANK (degree or grid size)=52024.060—
PowerNetdegree n=322019.110—
ReLU2024.060—
PowerNetdegree n=242019.110.0001—
ChebNetdegree n=242019.110.0001—
Jacobi-rKANK (degree or grid size)=62024.060.0001—
Padé[K/5]-rKANK (degree or grid size)=62024.060.0003—
Padé[K/6]-rKANK (degree or grid size)=42024.060.0003—
PowerNetdegree n=162019.110.0004—
ChebNetdegree n=162019.110.0004—
Padé[K/4]-rKANK (degree or grid size)=62024.060.0006—
Padé[K/5]-rKANK (degree or grid size)=42024.060.0007—
Padé[K/4]-rKANK (degree or grid size)=22024.060.001—
Padé[K/5]-rKANK (degree or grid size)=52024.060.0018—
Padé[K/5]-rKANK (degree or grid size)=32024.060.0036—
PowerNetdegree n=82019.110.004—
ChebNetdegree n=82019.110.0041—
Padé[K/5]-rKANK (degree or grid size)=22024.060.0044—
KAN2024.060.0152—
PowerNetdegree n=362019.110.0478—
Padé[K/6]-rKANK (degree or grid size)=62024.060.0479—
FCNN2-1width=168, rank=–, depth=12, trained parameters=312985, total parameters=312985, training time=26.7s / 10 epochs2024.062.422.75
FCNN2-2width=240, rank=–, depth=12, trained parameters=637201, total parameters=637201, training time=29.3s / 10 epochs2024.063.281.39
MMNN2 (S1)width=789, rank=36, depth=12, trained parameters=313630, total parameters=637120, training time=30.3s / 10 epochs2024.064.611.55
MMNN2 (S2)width=789, rank=36, depth=12, trained parameters=637120, total parameters=637120, training time=35.8s / 10 epochs2024.066.176.05
PowerNetdegree n=402019.11178,000—
PowerNetdegree n=442019.1155,300,000,000—