.png)
Predictive ML model with a user application, generated by historical flight and weather data from 2019 to provide more accurate estimates of departure and arrival times for passengers and airline operations.
Using historical flight and weather data, build a machine learning model that predicts future flight delays and cancellations. The model predictions are supported by an interactive user application that reports the probability of an input flight being delayed, based on a variety of features that have impacted historical delays. Supporting our predictive flight delay dashboard, are interactive visualizations that analyze the historical 2019 flight data to provide further visual summary. Following future improvements and model optimization, our application aims to provide more accurate estimates of departure and arrival times for passengers and airline operations.
Project Presentation - Slide Deck
GitHub Repository - Flight Predict Application
This endpoint helps predict the likelihood of flight delays based on various inputs provided by the user. The user enters flight details into an HTML form, and this data is processed and used to predict the probability of a delay.
Users enter flight details such as origin, destination, airline, and departure time into a form on the /predict endpoint. JavaScript processes this data and sends it to the server using an AJAX connection.
The server receives the input and extracts necessary details like airport names, flight dates, and times. It uses those inputs to gather live weather data from the openweathermaps and weather.gov APIs as well as monthly flight statistics from the database. These details are then prepared for prediction by encoding and scaling the data appropriately. The processed data is fed into a machine learning model, which predicts the probability of a flight delay. This probability is then sent back to the web page and displayed to the user.
This endpoint handles both displaying the form (GET request) and processing the prediction (POST request). When a user submits the form, the server processes the input, queries additional data, and returns the delay probability.
Several helper functions are used to streamline data processing:
Route supporting historical visuals (static and dynamic) to further visualize the delay trends in 2019 and the data utilized to build model.
Internal routes to JSON data and external links to APIs and datasets used necessary to access in order to support the machine learning model, predictions and supporting visualizations
.png)
