PROBLEM TO BE SOLVED: To provide a machine learning device capable of appropriately determining a timing of starting transfer of a substrate and a transfer route thereof according to a state at that time in the apparatus. A state information acquisition unit that acquires state information including the position of a board in an apparatus and an elapsed time in each unit, and whether or not to take out a new board from a cassette in a certain state, and which processing unit. It has a prediction model that predicts the value for performing the action of transporting to, and performs the selected action with the action selection unit that selects one action based on the prediction model by inputting the acquired state information. The instruction signal transmission unit that transmits the instruction signal and the operation result acquisition unit that acquires the operation result including the number of processed sheets and the waiting time, and the acquisition so that the larger the number of processed sheets and the shorter the waiting time, the larger the reward. It is provided with a prediction model update unit that calculates a reward based on the operation result and updates the prediction model based on the reward. [Selection diagram] Fig. 5
展开▼