Bicycle demand forecasting from O'Reilly Media webcast and report; Data Science in the Cloud with Azure ML and R

This experiment applies several non-linear regression methods to forecast bicycle rental demand using Microsoft's Azure Machine Learning cloud service and R. You can view an O'Reilly media [webcast](http://www.oreilly.com/pub/e/3292) discussing this experiment, and read the companion [O'Reilly media report](http://radar.oreilly.com/2015/01/getting-started-with-data-science-in-the-cloud.html). The R code for this experiment can be downloaded from the [Git repo](https://github.com/Quantia-Analytics/AzureML-Regression-Example). The data set can be found in Azure ML Studio, or you can load it from the .csv file provided [here](https://archive.ics.uci.edu/ml/datasets/Bike+Sharing+Dataset). Reference for these data is; Fanaee-T, Hadi, and Gama, Joao, 'Event labeling combining ensemble detectors and background knowledge', Progress in Artificial Intelligence (2013): pp. 1-15, Springer Berlin Heidelberg. ## Background If you are unfamiliar with using R with Azure ML first read my [Quick Start Guide to R in Azure ML Studio](http://azure.microsoft.com/en-gb/documentation/articles/machine-learning-r-quickstart). The code is available in this [Git repo](https://github.com/Quantia-Analytics/AzureML-R-Quick-Start). Companion videos are available: * [Using R in Azure ML](https://www.youtube.com/watch?v=G0r6v2k49ys). * [Time series model with R in Azure ML](https://www.youtube.com/watch?v=q-PJ3p5C0kY). There is a [Git repo](https://github.com/Quantia-Analytics/AzureML-R-Quick-Start) containing all of the code for the Quick Start Guide. ## R object serialization This example uses serialization and unserialization of R model objects. My tutorial and code for serialization and unserialization of R objects in Azure ML is [here](https://github.com/Quantia-Analytics/AzureML-R-Serialization) . There is a companion video available on [YouTube](https://www.youtube.com/watch?v=vk9Ic1F9YTk&feature=youtu.be).