Ad

Friday, December 8, 2017

Median the undervalued summary statistics - technical interviews

Usually medians are quickly glossed over in college statistics classes and beginner coding courses. It's a simple concept : pick the middle number in a sorted dataset. Turns out that sorting is non trivial in large datasets so finding the median is harder in real life. Also know it depends on whether the dataset has even or odd number of data points. Recursively partition data by the median of subpartitions helps break sorted data into chunks that are easy to store and retrieve. It also makes binary search much easier.

A slightly challenging interview question is finding the median or two sorted arrays. The solution is nontrivial. 

No comments:

Post a Comment

Codecademy - Machine Learning Fundamentals - Syllabus

Upgrade your skills with Codecademy's Pro Intensive, Machine Learning Fundamentals. Each unit will cover conceptual and syntax lessons ...