Gene expression

GCM

Microarrays measure gene expression in human tumor samples spanning several cancer types. This benchmark predicts which of fourteen tumor classes a sample belongs to, using the class labels supplied with the Global Cancer Map dataset.

ClassificationLicence: PublicOpenML-1106Target: classopenml ↗

At this operating point

Model comparison

p=10,000, n=100 · Higher Macro-F1 is better

Random Forest is marked with a diamond. Models are ordered by the selected metric.

Across sample budgets

Learning curves

Models

All metric values at this operating point

Model	Macro-F1	MCC	Balanced accuracy	ROC-AUC
Kumo Tabular (M)	0.7161	0.7817	0.7476	0.9856
TabPFN 3.5	0.6839	0.7537	0.7119	0.9652
Causilo	0.6511	0.7195	0.6833	0.9621
Logistic Regression	0.6385	0.7038	0.659	0.9668
Logistic Regression (tuned)	0.6312	0.6849	0.6448	0.9559
RealMLP	0.6303	0.6913	0.6452	0.9566
Extra Trees (tuned)	0.5714	0.6701	0.619	0.9503
MLP (tuned)	0.5422	0.6229	0.5733	0.937
Random Forest (tuned)	0.5323	0.6208	0.5667	0.9502
Random Forest	0.5285	0.6092	0.5571	0.9292
TabM	0.5046	0.5754	0.5229	0.9344
XGBoost (tuned)	0.4953	0.5851	0.5381	0.9288
XGBoost	0.4808	0.5606	0.5095	0.9084
Extra Trees	0.4774	0.5631	0.4976	0.9142
TabDPT	0.4486	0.5565	0.4871	0.9311
MLP	0.4467	0.5603	0.4895	0.9254
AutoGluon	0.4442	0.5423	0.4717	0.9277
LimiX2	0.01948	0	0.07143	0.5
MITRA	0.01948	0	0.07143	0.5
RealTabPFN v2	0.01948	0	0.07143	0.5
RealTabPFN 2.5	0.01948	0	0.07143	0.5
TabFM	0.01948	0	0.07143	0.5
TabPFN Wide 5k (ne3)	0.01948	0	0.07143	0.5