Calculation of probability of X2 test being an anomalous data

Course Q&A Machine Learning Specialization Unsupervised Learning, Recommenders, Reinforcement
1

Course 3- Week 1

Description:

Hello, I’m wondering if the probability of X2 test being an anomalous data should be p(X2test) = 0.00635 given feature x1 being 7 and feature x2 being 1, but the professor (Andrew Ng) got p(X2test) = 0.0021. I double-checked everything yet I can’t seem to find the leak in my calculations. I’m referring to the “Anomaly detection algorithm” video example.

Thanks.

2

Hi there!

The difference comes from the value of the second feature. If you calculate it with x1 = 7 and x2 = 0.5 (as mentioned in the video), you get 0.0021, which matches the video’s result.

Let me know if you have any other questions!