Abstract: Bike rental predictions forecasts the demand for bikes rentals in dependency of weather conditions like the temperature and calendric information e.g. holidays. To make predictions machine learning is used. The dataset contains some time features which we translate to indicator variables, as well as weather information for that day. Generally in bike rental systems it is very important that the administrators should know how many bikes are to be placed in each station, knowing this count enables them in arranging the required number of bikes at the stations and decide whether a particular station needs to have extra number of bikes or not. So in this research work we study prediction associations to enhance their administrations and items in view of clients' input.
Keywords: Random Forest, Neural Network, Prediction, Result Analysis
| DOI: 10.17148/IJIREEICE.2019.7214