Details of passengers' flights and their questionnaire responses are used to predict their overall satisfaction (green) or dissatisfaction (red). Initial values are the means of the data sample taken from Kaggle. If the predicted risk appears inconsistent with the final step in the decision sequence, please refer to the ToC documentation that explains how predictions are calculated for leaf nodes that fail to discriminate between the study classes.
Analysis of individual situations may indicate suitable remedies to improve passenger perceptions. Bulk changes to large data samples (perhaps increasing the in-flight wifi score) can indicate the effect on general traveller satisfaction on changes to in-flight services. The costs or savings of changes can then be measured against the potential impact on passengers.
A ToC model could support cost-benefit analysis to determine cost-effective enhancements for travel. ToC will not provide definitive answers, but can inform management decision making on investment or cost-saving.
This model is purely for illustration. Performance has been validated against data from the sample downloaded, but there has been no evaluation using current real-world data. Data codesets are unexplained to restrict its use.