Qlearning_DMS: Dynamic model selection using Q-learning for load forecasting...

In fengcong1992/QLearningForecast: Reinforced Load Forecasting via Q-Learning Dynamic Model Selection

Dynamic model selection using Q-learning for load forecasting Author: Cong Feng Reference: 1. Feng, C. and Zhang, J., 2019, February. Reinforcement learning based dynamic model selection for short-term load forecasting. In 2019 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT) (pp. 1-5). IEEE. 2. Feng, C., Sun, M. and Zhang, J., 2019. Reinforced Deterministic and Probabilistic Load Forecasting via $ Q $-Learning Dynamic Model Selection. IEEE Transactions on Smart Grid, 11(2), pp.1377-1386. This function performs dynamic model selection using Q-learning for load forecasting

Description

Dynamic model selection using Q-learning for load forecasting Author: Cong Feng Reference: 1. Feng, C. and Zhang, J., 2019, February. Reinforcement learning based dynamic model selection for short-term load forecasting. In 2019 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT) (pp. 1-5). IEEE. 2. Feng, C., Sun, M. and Zhang, J., 2019. Reinforced Deterministic and Probabilistic Load Forecasting via $ Q $-Learning Dynamic Model Selection. IEEE Transactions on Smart Grid, 11(2), pp.1377-1386. This function performs dynamic model selection using Q-learning for load forecasting

Usage

Qlearning_DMS(
  QMS_no_models,
  QMS_no_episodes,
  QMS_no_hour,
  QMS_init_md,
  QMS_reward_selection,
  QMS_alpha,
  QMS_gamma,
  df_learn,
  df_select
)

Arguments

QMS_no_models	state and action space dimension
QMS_no_episodes	number of episodes
QMS_no_hour	update frequency
QMS_init_md	starting state
QMS_reward_selection	reward strategy
QMS_alpha	learning rate
QMS_gamma	discount factor
df_learn	training data frame
df_select	selecting data frame

Value

the Q-learning forecasting results

References

Feng, C. and Zhang, J., 2019, February. Reinforcement learning based dynamic model selection for short-term load forecasting. In 2019 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT) (pp. 1-5). IEEE.

fengcong1992/QLearningForecast documentation built on Oct. 11, 2022, 1:47 p.m.