Subscribe to DSC Newsletter

The ability to recognize objects and their relationships is at the core of intelligent behavior. This, in turn, depend on one’s ability of perceiving similarity or dissimilarity between objects, be physical or abstract ones. Hence, if we are interested to make computers behave with any degree of intelligence, we have to write programs that can work with relevant representation of objects and means to compute their similarities or lack thereof, i.e., dissimilarity (obviously, they are two faces of the same coin).

The white paper: Measuring Similarity between Objects 

Views: 1430

Tags: distance, feature, similarity, space


You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Colin Goldberg on October 19, 2016 at 7:15am

Congratulations on an article written with a clarity that I appreciate. As someone whose math is a bit rusty, I will need to reread it - probably a number of times - so I can truly understand it.

One thing that was missing that I would like to see is an example with real number calculations - these would serve to clarify specific points even more as there is a lot of material to go through. Ideally I would like to see the example in a language such as python that I could reproduce on my computer.

A question arises: "What is the relationship between predictability and similarity?". As a novice in the area of data science, nevertheless I believe exploration into this topic would facilitate my coming to grips with this subject. I ask this in the context of my looking at prediction models and algorithms (such as XGBoost) and wondering what is the overlap - if any - between them. My gut feeling is that similarity goes further than predictability - but this is only from my perspective as a novice. As a pragmatist, a python example or two would be very helpful.

I look forward to a greater understanding of this subject. Any pointers or links would be much appreciated.


Colin Goldberg

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service