This repository contains the solution to the VUELING TECH HACK.
My name is Guillermo Seoane and I'm a #DataScience student at the IT Academy de Barcelona Activa.
At the last Annual General Assembly of IATA, the zero net CO₂ emissions in 2050 (aviation sector) resolution finally got approved. That lets us be one step closer to the Paris Agreement of 2015, accomplishing not exceeding 1.5 ºC the Earth's temperature.
In the repository you can find:
- Exploratory_analysis of the data,that allows you to understand it better and find insights (ex: flights per month, most used routes, etc.) based on "Mapping Destination Countries". Download 'Mapping Destination Countries'
- Model: model used for prediction.
- A presentation.pdf explaining how I have reach the solution
- Date: Flight date.
- Origin_Country : Country of origin.
- Origin_Continent : Continent of origin.
- Destination_Country: Country of destination. Target variable.
- Destination_Continent: Destination continent.
- Total_flights: Total number of flights.
- Total_seats: Total number of seats.
- Total_ASKs: (Available Seat Kilometer). Total seat numbers available by the total number of km these seats have flown.
The data has been split into two groups:
- The training set has been used to build the machine learning models. Download 'train.csv'
- The test set has been used to see how well our model performs on unseen data. For the test set, the Destination Country variable is not provided. Download 'test.csv'
The metric used to evaluate the predictive algorithm has been F1 Score Macro.
- Model: RandomForestClassifier
- F1-Score: 0.97
- Download 'JSON'