Browsed by
Author: Manojit Nandi

Manojit Nandi is data scientist at STEALTHbits Technologies. He holds a Bachelor of Science in Decision Sciences, focused on machine learning and mathematical algorithms, from Carnegie Mellon University. He has given talks on machine learning and data science at tech conferences, such as SIGKDD and PyData.
Local Outlier Factor – Part 3

Local Outlier Factor – Part 3

Part 3 Part 3 – Local Outlier Factor scores & Conclusion In Part 2, I demonstrated how to compute the local reachability density for each data point. The local reachability density grants us insight as to how “isolated” a data point is. In this third and final blog post, I will compare each point’s local reachability density to its neighbors to compute the local outlier factor score for each point. Based on the computed local outlier factor score, we can…

Read More Read More

Local Outlier Factor – Part 2

Local Outlier Factor – Part 2

PART 2 Part 2 – Reachability-Distance & Local Reachability Density In Part 1, I motivated the need for a density-based approach to outlier detection, and then I talked you through the first step of using the K-Distance as a framework for quantifying how “distant” a data point is from its neighbors. Today, I will walk you through the next two steps that will allow us to compute the local density of each data point. Reachability-Distance Now that we have the…

Read More Read More

Local Outlier Factor – Part 1

Local Outlier Factor – Part 1

User Behavior Analytics Finding Anomalous Users through the Local Outlier Factor Algorithm Part 1 – Motivation and K-Distance How do you identify users who are behaving anomalously? One way to tackle this problem is to define a set of rules that all user activity should conform to. If a user’s behavior breaks one or several of these rules, then we flag that user as behaving anomalously. There are several problems with this approach: 1. You need a good, formal definition…

Read More Read More

Start a Free StealthAUDIT® Trial!

No risk. No obligation.