Prediction of Capital bikeshare demand

Goal: In this project, solve a regression task on rental bikes.

This project was completed in the third week of The Data Science bootcamp at Spiced Academy

Description

I used the Capital Bikeshare dataset which contains daily records on the number of bikes rented as well as weather conditions.

step 1: Exploratory data analysis to detect patterns in the bike rental demand.

step 2: Applied Feature enginnering on the data to improve model.

step 3: Using different regression models to predict how many rental bikes are needed at a certain time, based on information about weather conditions and daily patterns

step 4: Checking the model by cross validation

I have got following results:

• Linear Regression (RMSLE: 0.60)
• Random Forest Regressor (RMSLE score: 0.16)
• Gradient Boosting Regressor (RMSLE: 0.33)