Overview
This Bicycle Rental Algorithm was made with the programming language Python. It took actual bicycle data from a
city and the goal was to create a predictive algorithm based on factors like the weather and the day of the week.
Skills
Programming
Data
Algorithms
Process
I started off with creating buckets where the actual data points would fall in to. From there, I had the algorithm
look at the trends and correlations. I tested it using multiple combinations of number of buckets and factors. To test
the algorithm, I had it randomly select actual data points to see if it could predict the number of bikes rentals for
that day within relative error.
The results vary but a typical outcome is like the picture below. About 60% accuracy with about 15% relative error.
