If you're in the Philadelphia area and want to hack on some data with other R users this evening, come to the IntegriChain/Revolution Analytics Meetup at Drexel University. There will be an short lecture to introduce the problem, followed by a hacking session (virtual machines with Revolution R software and data will be provided). And of course, plenty of food and drink will be provided to keep you going through the evening. Register here (free!), and full details are below.
IntegriChain/Revolution Analytics Meetup
May 14th, 2014, 7PM
What: The Challenge
Manufacturers find themselves in the constant struggle to accurately forecast product sales. By producing too much they waste capital by sitting on unused stock, alternatively producing too little they risk not having enough when an order shipment comes in. Being able to recognize purchasing patterns can allow manufacturers to tighten their supply chains and save lots of money. This presents an interesting problem to which big data and statistical analysis tools (such as R) can help provide an answer.
Participants in this challenge will receive actual sales data from a manufacturer (that has had any identifying information redacted) and are asked to predict the next period of sales.
The data provided will have the following fields:
Trade Partner – the wholesaler (or retailer) to whom the shipments are being made
Package – the product or SKU being sold
Date – sales is reported at the day level
QS – the quantity sold in appropriate units
Participants are asked to work together in teams to accomplish this task and are encouraged to experiment with different models and levels of aggregation. The resulting forecasts will be ranked by the sum of square error difference between the predicted and the known sales by package for each day in the “future” period.
When: May 14th, 7-10 PM EST
Where: Drexel CCI, Rush Building (014) – 30 North 33rd Street, Philadelphia, PA
What to Bring: Your laptop and your brain
What is Provided: Actual data, Revolution R Enterprise virtual machines, food and beverages, help
Registration: Working With R in Hadoop Lecture/Mini-Codefest