Criteo_x4#
Note that we have set two evaluation protocols Criteo_x4_001
and Criteo_x4_002
for this benchmark, which vary in the settings of rare feature filtering and embedding dimensions.
Criteo_x4_001
Please use the following evaluation settings for this benchmark:
Dataset split: Criteo_x4
Rare features filtering: min_categr_count=10
Embedding size: 16
Criteo_x4_002
Please use the following evaluation settings for this benchmark:
Dataset split: Criteo_x4
Rare features filtering: min_categr_count=2
Embedding size: 40
🔥 See the benchmarking results on Criteo_x4_001:
from plots import show_table, show_plot
show_plot("criteo_x4_001.csv")
show_table("criteo_x4_001.csv")
🔥 See the benchmarking results on Criteo_x4_002:
from plots import show_table, show_plot
show_plot("criteo_x4_002.csv")
show_table("criteo_x4_002.csv")
Rank | Year | Publication | Model | AUC | Logloss | Running Steps | Contributor |
---|---|---|---|---|---|---|---|
Loading... (need help?) |