project Details

This project applies skills in machine learning, deep learning, and data visualization to optimize flight flows in the United States, with the goal of reducing delays and improving air traffic efficiency.

Results:

This project received a grade of 18 out of 20, delivering robust and comprehensive analyses that effectively address the questions posed.

View Project Code

View on GitHub

Flight Data Analysis in the United States

We have data on the details of arrivals and departures of commercial flights for major airlines in the United States, covering the years 2000 to 2002. The goal is to perform the necessary analyses to answer the following questions:

Delay Analysis:

What are the best times of day and days of the week to minimize flight delays for each year?

Impact of Aircraft Age:

Do older aircraft experience more delays year over year?

Modeling Flight Diversions:

For each year, fit a logistic regression model to predict the probability of flight diversion within the United States. Use as many variables as possible, including the following attributes:
  • Departure date
  • Scheduled departure and arrival times
  • Coordinates and distances between departure and arrival airports
  • Airline carrier

Visualize the model coefficients for each year.

Delay Prediction Model:

Develop a machine learning or deep learning architecture of your choice to build a model capable of predicting arrival delays.