Skip to main content
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: Med Phys. 2021 Jan 11:10.1002/mp.14712. doi: 10.1002/mp.14712

Table 2.

Comparison of performance using rules only, DRL and KgDRL on testing dataset.

Initial Rules DRL KgDRL
Number of training episodes -- -- 8 100 8
Training time (hours) -- -- 13 172 13
Average PlanIQ score (± standard deviation) 4.97 (±2.02) 7.81 (±1.59) 5.87 (±2.37) 8.43 (±0.48) 8.82 (±0.29)