Abstract: Systems and methods for implementing a reinforcement machine learning framework for dynamic demand forecasting. A method includes generating estimated booking data for an initial time with a demand model trained using a training set of historical booking data. A variance is detected between the estimated booking data and transient booking data observed at the initial time that exceeds a defined threshold. In response to detecting the variance, a reinforcement learning service is activated. An updated training set including enhanced booking data observed at a subsequent time is created after activating the reinforcement learning service. A parameter of the demand model is updated by training the demand model using the updated training set.

Abstract: Systems and methods for implementing a machine learning framework for demand shock detection for dynamic demand forecasting. A method includes generating predicted booking observations with a demand model trained using a training set of historical booking data. Transient booking observations are obtained from an active database. An observed likelihood score is computed from the transient booking observations based on the demand model trained on the historical booking data. A demand shock threshold is computed based on the statistical relationship between a time to detection of the demand shock event and at least one shock detection criterion. An occurrence of a demand shock event is determined by comparing the observed likelihood score to the demand shock threshold.

Abstract: Methods of reinforcement learning for a resource management agent. Responsive to generated actions, corresponding observations are received. Each observation comprises a transition in a state associated with an inventory and an associated reward in the form of revenues generated from perishable resource sales. A randomized batch of observations is periodically sampled according to a prioritized replay sampling algorithm. A probability distribution for selection of observations within the batch is progressively adapted. Each batch of observations is used to update weight parameters of a neural network that comprises an approximator of the resource management agent, such that when provided with an input inventory state and an input action, an output of the neural network more closely approximates a true value of generating the input action while in the input inventory state. The neural network may be used to select each generated action depending upon a corresponding state associated with the inventory.