Gene expression

TCGA Head & Neck (HNSC, RNA-seq)

RNA sequencing measures gene expression in tissue samples from The Cancer Genome Atlas head and neck squamous cell carcinoma project. The benchmark uses normalized expression values to predict whether each sample is labelled as tumor or normal tissue.

Binary classificationLicence: Not specifiedTCGA-TCGA-HNSC_Gene-Expression-QuantificationTarget: sample_typetcga ↗

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
RealTabPFN 2.5	0.9745	0.9512	0.9657	0.9947
RealTabPFN v2	0.9744	0.9512	0.9759	0.9968
TabPFN 3.5	0.961	0.9259	0.9435	0.9962
LimiX2	0.9558	0.9163	0.9527	0.9905
TabPFN Wide 5k (ne3)	0.9539	0.9112	0.9425	0.9977
TabM	0.9501	0.9041	0.9416	0.9979
Kumo Tabular (M)	0.9396	0.8799	0.9498	0.9898
GeneICL	0.9215	0.8536	0.9174	0.9941
RealMLP	0.9193	0.8472	0.8944	0.9957
MLP	0.8959	0.8164	0.8486	0.9945
TabDPT	0.8941	0.7931	0.8794	0.9608
MITRA	0.8708	0.7663	0.8286	0.9766
Logistic Regression	0.8707	0.7701	0.8083	0.9932
Causilo	0.8433	0.7292	0.7847	0.9934
Random Forest	0.8301	0.7057	0.7625	0.991
AutoGluon	0.8093	0.6593	0.8508	0.9649
TabFM	0.7882	0.6435	0.7361	0.9939
MLP (tuned)	0.7569	0.5357	0.7444	0.7996
Logistic Regression (tuned)	0.7143	0.4628	0.6861	0.7998
XGBoost	0.6667	0.3885	0.6361	0.9552
Extra Trees	0.601	0.3525	0.5681	0.9918
Random Forest (tuned)	0.5869	0.2447	0.5713	0.7948
Extra Trees (tuned)	0.5483	0.1388	0.55	0.7956
XGBoost (tuned)	0.517	0.09123	0.5222	0.7484