Summary

In this chapter, we carried out machine learning (ML) data analysis tasks on flight performance data. One such task is the implementation of a regression model fitted on a training subset of data. Given a new or unknown flight with delayed departure data, this model was able to predict whether the flight under investigation made up for time lost and arrived at the destination on time. One important takeaway from this ML exercise is this—the origin to destination distance contributed most toward predicting time gained. Carrier delays contributed least toward a prediction. A longer flight, it turns out, is able to gain more time. 

This chapter provided the foundation to build more sophisticated models. A model with more predictor variables (for example, taking into account, the weather and security delays) could yield deeper, sharper predictions. That said, this chapter hopefully opens a window for opportunity readers to understand how flight performance insights could help travelers...