Active learning is an effective method to train a model efficiently. It lets the model decide where in the parameter space it wants to learn next. By using only the necessary samples, an active learning strategy can effectively cut down the number of training samples, thus significantly reducing the computational burden of labeling training samples and improving the overall training efficiency.
In the notebook, we will walk through a case study, where an active learning strategy is employed to train a Gaussian Process model adaptively.
You can find the companion blog here: