A SAS data scientist submits the following PROC HPNEURAL code and gets a 'CONVERGENCE NOT ACHIEVED' warning. Which adjustment is most likely to resolve convergence issues?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because convergence failures in PROC HPNEURAL are commonly caused by: (1) insufficient iterations (increase MAXITER=), (2) poorly scaled inputs causing erratic gradient updates (normalize input variables), and (3) a learning rate that is too large causing oscillation (reduce learning rate). A does the opposite — decreasing MAXITER reduces opportunities for convergence and increasing the learning rate exacerbates oscillation.
Full explanation below image
Full Explanation
B is correct because convergence failures in PROC HPNEURAL are commonly caused by: (1) insufficient iterations (increase MAXITER=), (2) poorly scaled inputs causing erratic gradient updates (normalize input variables), and (3) a learning rate that is too large causing oscillation (reduce learning rate). A does the opposite — decreasing MAXITER reduces opportunities for convergence and increasing the learning rate exacerbates oscillation. C adds complexity which typically worsens convergence, not improves it. D may help marginally in some cases but addresses activation choice, not the primary convergence factors.