run loci taking a very long time

Cloudy10 / loci

Implementation of the LOCI outlier detection algorithm in Python, based on the paper Papadimitriou, S., Kitagawa, H., Gibbons, P.B. and Faloutsos, C., 2003, March, Loci: Fast outlier detection using the local correlation integral.

MIT License

3 stars 1 forks source link

run loci taking a very long time #1

Open Sarvind1 opened 5 years ago

Sarvind1 commented 5 years ago

I don't know why this is happening but in your implementation as well as pyod's implementations the training time(fit or run_loci) take too much time and memory for 10,000X30 data, please try to run it for credit card fraud detection kaggle dataset

mavroudo commented 3 years ago

There are some bad choices in the implementation, for example, there is no use of the sorted critical values. But all things considered, this is an O(n^3*d) algorithm and can take along time. You might want to look for aLoci if still interested

Avv22 commented 3 years ago

@mavroudo. Thank you. How to find aLoci?