csle_system_identification.environment_evaluations package

Submodules

csle_system_identification.environment_evaluations.stopping_game_emulation_eval module

class csle_system_identification.environment_evaluations.stopping_game_emulation_eval.StoppingGameEmulationEval[source]

Bases: object

Utility class for performing emulation evaluations of the stopping game

static emulation_evaluation(env: gym_csle_stopping_game.envs.stopping_game_env.StoppingGameEnv, n_episodes: int, intrusion_seq: List[csle_common.dao.emulation_action.attacker.emulation_attacker_action.EmulationAttackerAction], defender_policy: csle_common.dao.training.policy.Policy, attacker_policy: csle_common.dao.training.policy.Policy, emulation_env_config: csle_common.dao.emulation_config.emulation_env_config.EmulationEnvConfig, simulation_env_config: csle_common.dao.simulation_config.simulation_env_config.SimulationEnvConfig) → List[csle_common.dao.emulation_config.emulation_simulation_trace.EmulationSimulationTrace][source]

Utility function for evaluating a strategy profile in the emulation environment

Parameters

  • env – the environment to use for evaluation

  • n_episodes – the number of evaluation episodes

  • intrusion_seq – the intrusion sequence for the evaluation (sequence of attacker actions)

  • defender_policy – the defender policy for the evaluation

  • attacker_policy – the attacker policy for the evaluation

  • emulation_env_config – configuration of the emulation environment for the evaluation

  • simulation_env_config – configuration of the simulation environment for the evaluation

Returns

traces with the evaluation results

csle_system_identification.environment_evaluations.stopping_game_pomdp_defender_eval module

class csle_system_identification.environment_evaluations.stopping_game_pomdp_defender_eval.StoppingGamePomdpDefenderEval[source]

Bases: object

Utility class for performing emulation evaluations of the defender POMDP

static emulation_evaluation(env: gym_csle_stopping_game.envs.stopping_game_pomdp_defender_env.StoppingGamePomdpDefenderEnv, n_episodes: int, intrusion_seq: List[csle_common.dao.emulation_action.attacker.emulation_attacker_action.EmulationAttackerAction], defender_policy: csle_common.dao.training.policy.Policy, emulation_env_config: csle_common.dao.emulation_config.emulation_env_config.EmulationEnvConfig, simulation_env_config: csle_common.dao.simulation_config.simulation_env_config.SimulationEnvConfig) → List[csle_common.dao.emulation_config.emulation_simulation_trace.EmulationSimulationTrace][source]

Utility function for evaluating policies in the emulation environment

Parameters

  • env – the environment to use for evaluation

  • n_episodes – the number of episodes to use for evaluation

  • intrusion_seq – the sequence of intrusion actions to use for evaluation

  • defender_policy – the defender policy to use for evaluation

  • emulation_env_config – the configuration of the emulation environment to use for evaluation

  • simulation_env_config – the configuration of the simulation environment to use for evaluation

Returns

traces with the evaluation results

Module contents