Skip to main content
. 2021 Mar 17;18(176):20200987. doi: 10.1098/rsif.2020.0987

Figure 4.

Figure 4.

EQL pipeline. Step 1: Generate averaged ABM output; Step 2: Estimate the temporal derivative of the ABM output; Step 3: Library construction; Step 4: Equation inference. At Step 4, one can either perform (a) regression or (b) sparse linear regression to learn an equation for the ABM output. We will consider both the Lasso and Greedy sparse regression algorithms to perform EQL.