2 — ML Algo Interview Series: K-Means

VaticAI
5 min readApr 15, 2022

One of the foundational clustering algorithms, it’s one of the best contributions of the signal processing world to ML applications. Invented as a method of vector quantization (partition n observations into k clusters). Considered a simple algorithm, K-means is often overlooked in ML interviews. Nonetheless, candidates should have a good idea of a foundational and versatile clustering algorithm as it comes in handy in several real-world applications.

This post is the 2nd part of our 20-part series covering one applied ML algorithm every week. It’s meant to quickly brush up on concepts and validate some forgotten pointers.

Yes or No Format

1.1. Is K-Nearest neighbor a supervised or unsupervised learning algorithm (Trick question)

1.2. Is K-means a supervised or unsupervised learning algorithm (Trick question)

1.3. k stands for a number of clusters in K-Means?

1.4. Does centroid initialization affect K means Algorithm?

1.5. Does k-means work if clusters are of different sizes and variances?

1.6. Is k-means sensitive to outliers?

One-Line answers

2.1. Why do you prefer Euclidean distance over Manhattan distance in the K means…

--

--