Trains a model for predicting the performance gain of a bootstrapping process prior to actually applying it using two features; the baseline classifier's maturity (i.e. how close is the current hyperplane to the optimal) and the oracle's reliability (i.e. how reliable is the oracle in providing the correct labels of new training data). The code can be downloaded from github or download directly from here.
Elisavet Chatzilari, Spiros Nikolopoulos, Yiannis Kompatsiaris and Josef Kittler, “PERFORMANCE PREDICTION OF BOOTSTRAPPING FOR IMAGE CLASSIFICATION”
Figure 1: Actual cumulative performance gain with respect to the number of classifiers enhanced, using the proposed regression model to predict the expected performance gain of each classifier. Comparison with three baselines; a) Random, b) Upper and c) Lower baseline.
# of enhanced classifiers