The project is a Jupyter Notebook that contains a complete Data science methodology to analyse a simple yet interesting problem: route grades.
Climbing is a sportive discipline where participants must... climb, either indoor or outdoor. Here, we tackle indoor climbing: a wall contains holds, and the climber must only use some of them. A difficulty ("grade") is attributed following the pattern: (number in range 5-9)(letter in range a-c)(nothing or a + symbol) (for example: 5c+ or 8b).
The goal is to predict the grade of the route only using its physical properties, and not the feeling of the climber. We use Deep Learning algorithms inside a global Data science method to solve this problem.
Everything is printed in the PDF file
- Data from Moonboard