BMC Bioinformatics

Table 4 Process-based models vs ML models (3x1 combinations)

From: Predicting rice blast disease: machine learning versus process-based models

Dataset	M5RULES			RNN			YOSHINO			WARM
Dataset	r	r²	%MAE*	r	r²	%MAE*	r	r²	%MAE*	r	r²	%MAE*
K2016	0.76	0.58	0.52	0.81	0.66	0.49	0.84	0.71	0.23	0.31	0.10	0.77
S2016	0.77	0.60	0.88	0.75	0.56	0.95	0.47	0.22	0.50	N/A**
K2015	0.29	0.09	0.49	0.72	0.52	0.58	0.34	0.12	0.75	0.78	0.61	0.76
P2015	0.57	0.32	0.62	0.53	0.28	0.99	N/A**			0.69	0.48	0.92
Avg.	0.59	0.39	0.63	0.70	0.50	0.75	0.55	0.35	0.49	0.59	0.40	0.82

*Mean absolute error, **Results are not available for these locations

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com