Fig. 2: FastMapSVM’s sensitivity to train data size and dimensionality of Euclidean embedding. | Communications Engineering

Fig. 2: FastMapSVM’s sensitivity to train data size and dimensionality of Euclidean embedding.

From: Classifying seismograms using the FastMap algorithm and support-vector machines

Fig. 2: FastMapSVM’s sensitivity to train data size and dimensionality of Euclidean embedding.

Shows the F1 score for varying train data size and dimensionality of Euclidean embedding. a, b These results for train data size and dimensionality of Euclidean embedding on the horizontal axis, respectively. Error bars represent the standard deviation of the measurements over 20 trials. Marker and line colors in (a) represent the number of dimensions of the embedding. Marker and line colors in (b) represent the number of instances in the train data set.

Back to article page