My project contrasts K-Nearest Neighbors and Random Forrest Regressors on Real World data

In many areas, rental bikes have been launched to improve accessibility ease. It is important to have the rented bike ready and open to the public at the appropriate time, as this reduces the amount of time people have to wait. Eventually, ensuring a steady supply of rented bikes for the area becomes a big concern. The most important aspect is predicting the number of rental bikes required at each hour in order to maintain a steady supply. In this project, we discuss the ways in which we can predict the number of bikes needed for the particular day based on the provided data set. These type of prediction systems enable users to borrow a bike from a specific location and return it to a different location. Hence, we use machine learning to predict the number of rental bikes that are needed on a particular day


In Machine Intelligence, there are many ways in which we can predict the number of bikes that might be needed in a particular day. One of the methods used was to examine the models for predicting hourly rental bike demand and investigate a function filtering method to exclude non-predictive parameters and rate features based on their prediction efficiency. The project was accomplished by using repeated cross validation to train five statistical regression models with their best hyper-parameters, and then evaluating their results. The other method just estimates the cumulative number of rented bikes in the entire bike sharing system. The various data in the data collection were used to manipulate and forecast the final number of rental bikes. Methods such as Ridge Linear Regression, Support Vector Machine for Regression, Random Forest Method for Regression and Gradient Boosted Regression Tree are used for the prediction of rental bikes.

Additional Info:

Feel free to dowload my code which is in I have also provided a copy of the testing and training data sets used. Lastly, I have also uploaded a copy of the short research paper that I wrote based on this project.


View Github