Model should be agnostic to the number of sensors that can be placed on the plate with heat

Heat Signature Zero
1

I have doing some research on leveraging an adjoint-based localization method (AdjointSparseLocator) to try and solve the backward heat equation to estimate source locations from sensor data. This doesn’t work where you have sources less than 4 hence I have resorted to using distance-weighted averaging. Maybe we can get clarification whether the private test dataset has the same number of sensors or more or less.

2

Hi @Adeptschneiderthedev and welcome to the Challenges community :waving_hand:

The parameter ranges, including the number of sensors and sources, will be relatively similar, but their distributions will differ.

ThinkOnward Team