Conclusions

Clustering emerges as a vital data analysis tool widely employed by data professionals to discern patterns within datasets. In the current study, both K-means and Hierarchical clustering, despite yielding distinct clusters, provided valuable insights. 

Upon evaluating the optimal clusters, it becomes evident that delays in Florida predominantly occur at high temperatures, specifically exceeding 75°F. While temperature appears to be a contributing factor, it is imperative to acknowledge that it cannot be solely attributed to delays. Subsequent investigations reveal additional influential factors in Florida, with humidity playing a significant role. Notably, during the summer season, characterized by elevated temperatures, Florida experiences frequent thunderstorms, resulting in low visibility, severe turbulence, and safety concerns. 

Further contributing to the complexity of delays in Florida is the fact that hot air, prevalent during high temperatures, is less dense than cold air. This phenomenon leads to decreased lift generated by an airplane's wings, potentially contributing to the clustering of delays in higher temperatures in the region. 

Moreover, the inclusion of additional features such as humidity and thunderstorm conditions can enhance clustering, providing a more nuanced understanding of the causes of delays. Future applications of clustering could extend beyond Florida to encompass all states in the U.S., allowing for the identification of patterns specific to each region. It is noteworthy that, for clustering geospatial data, alternative algorithms may prove more suitable than K-means, urging consideration of diverse clustering methodologies. 

In conclusion, clustering techniques, exemplified by K-means and Hierarchical clustering in this study, have proven instrumental in revealing patterns and associations within weather-related flight delays. The identified clusters, particularly highlighting the prevalence of delays in high temperatures in Florida, offer valuable insights into the multifaceted nature of flight disruptions. Looking ahead, there is potential for further research to enhance our understanding of optimal conditions and additional factors contributing to flight delays.