Fig. 7
From: Personalized tourism recommendation model based on temporal multilayer sequential neural network
Sensitivity analysis. Sensitivity analysis was conducted from two aspects: Impact of Time Interval Size and Influence of Track Length. (a) (filled blue square—Recall, filled orange triangle—Precision) illustrates the trend of the impact of different time intervals on the prediction results, identifying the optimal time interval value as 0.8 to 1.2 h. (b) (filled blue square—Recall, filled orange triangle—Precision) shows the influence of input and output trajectory lengths on the prediction results, concluding that both excessively long and excessively short trajectory lengths are detrimental to the model’s training.
