Today machine learning algorithms (e.g. KNN et. al) seem to be the obvious choice for any problems related to and manipulation of nearest data points.

Even though ML algorithms are appropriately suitable for this purpose, many of the problems can also be solved by using spatial triangulation concepts like Delaunay, Voronoi, etc. These concepts can be implemented as data structures and we have that as part of SciPy spatial library.

Using SciPy data structures reduced the complexities associated with using ML algorithms like training, etc, thereby reducing the time it takes to implement the same. It also requires less CPU and Memory Resources to do its work.

The lack of awareness about the spatial triangulation concepts and the availability of SciPy libraries for the same, developers, end up using ML algorithms as a one stop solution which may or may not be required.

Welcome to this talk on SciPy spatial data structures, where you’ll not only understand the algorithms used in modern day network infrastructure, but will also be able to see them in action online.

By the end of this talk, you’ll know enough to make an informed decision about using ML algorithms Vis-a-vis SciPy Spatial Data Structures

The presentation as well as the codes will be shared by GitHub Page post session.