We are quickly becoming big fans of Google Cloud Dataflow. See our previous posts on this topic here and here. We are excited to continue using this product and creating analytics solutions for our customers. In this post we are going to describe how RFM technique can be applied on large data sets – think 100’s …
Customer segmentation – RFM technique on big data using Google Cloud Dataflow Read More »
Consumers, customers, clients or users. Call them what you will, but businesses exist and grow only when they can serve them well and attract more of them. While there are several business strategies that one employs, one key methodology is to group or segment similar customers based on past purchasing behavior and geodemographic information. This helps …
This post analyzes cricket’s (http://en.wikipedia.org/wiki/Cricket) one day international games to identify batsmen who helped their teams in come from behind wins. This is an attempt to identify if some of the big names in cricket are truly worthy of their reputation. Once matchwinners for past games have been identified (train data), a statistical model based on …
We recently started a prototype to implement an end to end analytics solution using Google Cloud Dataflow. Google has done an excellent job documenting the various steps to get started as well as provided several examples to implement analytics solutions in Dataflow. Here’s a quick summary of what I had to do to get started. Download Google …
Google Cloud Dataflow has recently entered beta. It provides a powerful big data processing platform in the cloud. It’s an end to end solution that comprises compute resources, big data storage, data pipeline development and execution. Google has open sourced the Cloud Dataflow SDKs that allows users to define data processing jobs. Here’s a high level …
This blog post focuses on model evaluation. Accuracy and AUC (Area Under Curve) are two metrics we focus on since they are good for binary classifications. Accuracy deals with ones and zeros, meaning you either got the class label right or you didn’t based on a certain threshold cur off. The most natural threshold is 0.5 …
This blog focuses on feature selection and data exploration. AzureML provides an easy way to visualize data. Histograms and boxplots are two of the most common way to visualize data and understand data attributes. A histogram is a graphical representation of the distribution of numerical data. It is an estimate of the probability distribution of a continuous variable …
Here’s my first AzureML model: http://gallery.azureml.net/Details/6f058343b2224057a8ae259311f7ed31. The model uses a two class logistic regression algorithm for binary classification. This is based on Python based sample from learnds.com. It leverages various AzureML Studio components as well as custom R code for data cleansing and feature engineering. Here are the various steps I followed to get the model …
