Skip to main content
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: IEEE EMBS Int Conf Biomed Health Inform. 2016 Feb;2016:577–580. doi: 10.1109/BHI.2016.7455963

Figure 2. Stacked Generalization Improves Prediction Performance.

Figure 2

(A) Comparisons of stacked generalization to single data modalities for each cancer endpoint. Box plots indicate differences in prediction performance. Green/red-highlighted p-values indicate that the change in performance is positive/negative and statistically significant. (B) Stacked generalization out-performs the majority vote method of prediction modeling for the OV grade and KIRC grade endpoints. Both methods perform similarly for the KIRC survival endpoint. However, majority vote out-performs stacked generalization for the OV survival endpoint.