Last week, Dan Gilmore of Supply Chain Digest asked us to provide predictions for what would happen in the field of network design. Here is his full article with predictions in other areas.
Here our Top 3 predictions for 2016:
1. Modeling will become more collaborative. Modeling has always required data and feedback from many different people in the organization. Now, cloud-based technology now makes it easier to have different people in the organization work on the input files and explore the results.
2. Your team will learn the term data munging. Preparing data for a modeling exercise can involve working with 20-40 different data files and going through 100s of different steps. This is cumbersome in Excel and Access. Data munging refers to the process of going through those steps to prepare the data. But, more interesting, there are now tools(Alteryx and LLamasoft’s DataGuru to name two) that allow you to systematically execute and repeat these steps. These tools also make it easier to remember what you did when you return to work after a long weekend.
3. Machine learning will help you understand your data and automate the modeling building. When working with complex and detailed transaction data, machine learning algorithms will help you get deeper insights into your data. It will also help you automate the model building processing by cleaning the data, filling in missing gaps, and pointing out inconsistencies in the data.