Very slow execution for middle to large datasets

Question

Very slow execution for middle to large datasets

omarcr opened this issue 9 years ago · comments

Omar Costilla Reyes commented 9 years ago

Hi, is there any change in the code that you might recommend to speed up the execution of the algorithm?

sjqtentacles · Answer 1 · Wed Jul 29 2015 12:03:33 GMT+0800 (China Standard Time)

DTW in general is O(n^2), so that's expected. You could look into this https://en.wikipedia.org/wiki/Dynamic_time_warping#Fast_computation

Mark Regan · Answer 2 · Fri Sep 11 2015 00:59:18 GMT+0800 (China Standard Time)

Sorry for delay - missed this. As @Nixonite mentioned, DTW is O(N^2) and it is used within a non-optimized naive KNN algorithm which is O(dN). Combining these, you get a model that doesn't scale well.