Table 1 Percentage of risk-prone trajectories based on two decision criteria
From: Learning search polices from humans in a partially observable context
Greedy | GMM | Hybrid | Coastal | Human | |
|---|---|---|---|---|---|
Risk-prone (f) | 77% | 11% | 30% | 46% | 26% |
Risk-prone (r) | 78% | 12% | 24% | 45% | 7% |