Table 3.

Performance of rrBLUP (average r²) on the traits of four real data sets under the traditional encoding vs. the hybrid data-driven encoding.

Data Set	Traditional Encoding	Hybrid Data-driven Encoding	Improvement
Rice: Pericarp.color	0.433	0.504	16.4%
Rice: Protein.content	0.176	0.177	0.6%
Pig: Trait 2	0.237	0.239	0.8%
Pig: Trait 4	0.203	0.218	7.4%
QTLMAS: Trait 1	0.358	0.361	0.8%
QTLMAS: Trait 2	0.187	0.18	-3.7%
Maize: Flint 1 TASS	0.47	0.492	4.7%
Maize: Flint 2 DMC	0.301	0.308	2.3%
Maize: Flint 3 DM_Yield	0.057	0.068	19.3%
Maize: Dent 1 Tass	0.59	0.616	4.4%
Maize: Dent 2 DMC	0.562	0.58	3.2%
Maize: Dent 3 DM_Yield	0.321	0.349	8.7%