The partial dependence of a model on feature is : the average prediction with feature set to for every observation (Friedman, 2001). The individual conditional expectation (ICE) curves are the terms of that average, one curve per observation (Goldstein and co-authors, 2015). Accumulated local effects (ALE) average, within narrow intervals of feature , the change in prediction when moves across the interval for the observations that lie in it, and accumulate those changes (Apley and Zhu, 2020).
| correlation of and | |||
| root mean square error against the true effect of | 0 | 0.92 | 0.99 |
| partial dependence | 0.027 | 0.134 | 0.229 |
| accumulated local effects | 0.028 | 0.044 | 0.110 |
ml_explain.effects.ml_explain.curves.