Mastering .NET Machine Learning

Summary

In this chapter, we took a look at two common machine learning classifiers: k-Nearest Neighbors and Naïve Bayes. We saw them both in action with our AdventureWorks dataset to see if we can increase cross sales. We saw that k-NN had some limited success and Naïve Bayes was not useful. We then used our old friend logistic regression to help us narrow down a specific bike model that can be used to promote cross sales. Finally, we considered that since the data is ad hoc, we can't implement any real-time training on our website. We would want to periodically run this analysis to see if our original findings continued to hold.

In the next chapter, we are going to take off our software engineer hat and put on our data scientist hat to see if ...