Complete result settings for Netflix dataset (table columns are sortable)
The experiments are described in Rooshenas and Lowd, Discriminative Structure Learning of Arithmetic Circuits, AIStats 16
Standard Deviation | L1 Penalty | Train CLL | Validation CLL | Test CLL | Node# | Edge# | Feature# | Learning Time |
0.5 | 2 | -27.001669 | -27.230325 | -27.282087 | 578278 | 1334757 | 2574 | 35745.510855 |
0.5 | 0.5 | -27.004619 | -27.236546 | -27.288256 | 578278 | 1334757 | 2574 | 38537.004484 |
0.5 | 1 | -27.004239 | -27.235239 | -27.287175 | 578278 | 1334757 | 2574 | 38657.371185 |
0.5 | 0.1 | -27.004565 | -27.236472 | -27.288514 | 578278 | 1334757 | 2574 | 39461.787896 |
0.1 | 1 | -27.001579 | -27.232084 | -27.283631 | 578278 | 1334757 | 2574 | 44680.572520 |
0.1 | 0.5 | -27.004619 | -27.236546 | -27.288256 | 578278 | 1334757 | 2574 | 45613.233734 |
0.1 | 2 | -27.001669 | -27.230359 | -27.282079 | 578278 | 1334757 | 2574 | 48151.085921 |
0.1 | 0.1 | -27.004565 | -27.236472 | -27.288514 | 578278 | 1334757 | 2574 | 48442.486622 |