Anomaly detection with KNN

How do you know something is not right or it is far from the normal situation? Mathematically, if we can measure the distance between the new observation and the rest of the dataset (observed earlier), we can judge the closeness of this new data point to the historical dataset. In many applications, if we have fair confidence in the normality of the historical dataset, the low distance would show the normality of new observation. KNN is a simple way to measure the distance between each point and it's K neighbors.
🔴 Subscribe for more ML projects: www.youtube.co...
💻 Store sales anomaly detection • How to find anomalies ...
💻 Anomaly detection with isolation forest • Anomaly detection usin...

[@coeurblanc4999]

good video. suggest to turn up the volume. good content nonetheless. thanks

[@peymanrazmi5909]

Excellent Dr.Mohammad. Are these types of algorithms (KNN) considered as weak algorithms in ensemble learning? Please make the similar video and post for other algorithms.

[@michael_bryant]

Thanks, really helpful video

[@MikeSaintAntoine]

Great video!

[@ugurileri9936]

very helpful video; I want to ask one question about time series part; you have entered n_neighbours=5 why 5? What about if it is 2 or 3 or 4? If I use time series anomaly detection part for 4 - 5 sensors column data; what should I choose for n_neighbours parameter? again 5?

[@AIwithDrMo]

Thanks Ugur. n_neighbours depends on your application and we usually try different ones to see if the outputs makes sense for this specific project or not.

[@123hfksue3]

hello, where i can find your dataset?? can you share?

[@user-pi5qm6wu5k]

excellent!!

[@alwaaffa]

You can help me with a master’s thesis for my software part (coding) in Python?

[@AIwithDrMo]

Please fill out the following form for any specific questions, forms.gle/Jz4pkrNSGUqGhPug9

[@alwaaffa]

@@AIwithDrMo I can connect with you by email?