Complete result settings for WebKB dataset (table columns are sortable)
The experiments are described in Rooshenas and Lowd, Discriminative Structure Learning of Arithmetic Circuits, AIStats 16

Split Penalty Standard Deviation L1 Penalty Train LL Validation LL Test LL Node# Edge# Feature# Learning Time
10 0.1 2 -151.429662 -163.353407 -159.995830 259403 595612 4090 3455.588670
10 0.1 1 -151.318948 -163.254981 -159.910287 303025 699944 4160 4406.634089
10 0.1 0.1 -150.994158 -162.970802 -159.685810 290002 668969 4288 5698.654672
10 0.5 2 -150.553048 -162.568664 -159.294552 392432 913769 4388 7901.824739
10 0.5 1 -150.142017 -162.447047 -159.103252 396964 931557 4518 10419.382012
10 0.5 0.1 -149.669189 -162.100710 -158.817777 464914 1091752 4700 15040.924431
5 0.1 2 -149.353846 -161.475530 -158.324324 439156 1011031 5096 19926.391727
5 0.1 1 -148.893915 -161.305671 -158.224461 419177 964494 5314 22744.704279
5 0.1 0.1 -148.395816 -160.874242 -157.736417 529944 1231602 5678 42720.795450
5 0.5 2 -148.151551 -160.848921 -157.687448 603902 1408510 5616 43074.699649
5 0.5 1 -147.415174 -160.403409 -157.313682 784568 1833179 6098 71036.615792
2 0.5 1 -146.525988 -159.933088 -156.654906 141710 305631 7834 75093.626032
2 0.5 0.1 -146.237628 -159.907838 -156.711250 153784 331038 8218 75135.463673
2 0.1 0.1 -146.272960 -159.496258 -156.401111 276875 622897 7956 75657.759272
2 0.5 2 -146.383688 -159.537720 -156.291976 288441 656419 7492 75672.505030
2 0.1 2 -146.582254 -159.814804 -156.607956 602591 1396702 7088 75582.853658
2 0.1 1 -146.107219 -159.359885 -156.255311 596676 1376735 7684 75819.737647
5 0.5 0.1 -146.817744 -160.002226 -156.959195 793504 1863546 6510 76076.881555