Prepping Geo-Location Data for Machine Learning
In my last two posts, I described the techniques we used to create and enhance an ML data set for supervised learning. In my next post, we finally get to the really fun stuff – building models. But...
Ugly isn’t always a dirty word : The Measurement Minute #49 w. Gary Angel
The Measurement Minute #49 with Gary Angel : Ugly Isn't Always a Dirty Word A recent article in Slate magazine on the Trump campaigns brutally ugly emails not only brought a lump of nost...
Retiring the “Organizations that Do X are Y% more likely to be Z” Business Trope : The Measurement Minute #48 w. Gary Angel
The Measurement Minute #48 with Gary Angel : Retiring the "Organizations that Do X are Y% more likely to be Z" Business Trope The X,Y,Z trope is a stock in trade of Analyst firms. But ev...
Don’t be a Friggin Moron – Choosing a Cloud Provider for Retailers : The Measurement Minute #47 w. Gary Angel
The Measurement Minute #47 with Gary Angel : Don't be a Friggin Moron - Choosing a Cloud Provider for Retail Some IT decisions aren't really about IT. And if you're a retailer, choosing ...
Building a Data Factory for ML Learning
There’s a rich history around the idea of data augmentation for Machine Learning. If you have a library of images you want to train a model on, you can use simple image manipulation techniques to c...
Too Sexy to Be Useful? : The Measurement Minute #46 w. Gary Angel
The Measurement Minute #46 with Gary Angel : Too Sexy to be Useful - Data Science in the Enterprise Have enterprise's sacrificed too much in trying to make data science jobs as sexy as p...
The Analytics Nexus : The Measurement Minute #45 w. Gary Angel
The Measurement Minute #45 with Gary Angel : The Analytics Nexus Free. Online. Slack. Cutting-edge analytics. That's a pretty great formula for a digital analytics conference. ...
Operationalizing ML : The Measurement Minute #44 w. Gary Angel
The Measurement Minute #44 with Gary Angel : Operationalizing ML Like most things, we tend to focus on the fun and interesting part of machine learning - creating models. But in many res...
Supervision Required: Why all the data in the world may not be enough
We use a variety of sensors to track shopper movement in stores. And to do that more accurately, we re-engineered our processes to take advantage of machine learning. You can get the full skinny on t...