Figure 12From: Learning search polices from humans in a partially observable contextReduction of the uncertainty for the uniform, #1, #2 and #4 experiments. The expected value is reported. Top left: uniform initialisation, expected uncertainty for the greedy (red), GMM (blue), hybrid (green) and coastal (orange) search strategies. Top right: experiment #1. Bottom left: experiment #2. Bottom right: experiment #4.Back to article page