FIGURE 7.
Time in hours for implementing two sparse kernels (Arc_cosine_1 and Gaussian) with the maize data set as a function of the proportion of the compression level (0.5, 0.4, 0.3, 0.2, 0.1 and 0), which corresponds to using only the following proportions: 0.5, 0.6, 0.7, 0.8, 0.9 and 1 of the original lines (LinesProportion) for computing the sparse kernels since level of compression = 1 minus proportion of lines used to compute the sparse kernel. (A) corresponds to M4 and Arc_cosine_1 sparse kernel. (B) corresponds to M5 Arc_cosine_1 sparse kernel. (C) corresponds to M4 and Gaussian sparse kernel. (D) corresponds to M5 Gaussian sparse kernel. M4 denotes the random forest model and M5 denotes the Bayesian regression model.