How do I approach Coding & Algorithms interview questions?

Coding & Algorithms questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master coding & algorithms interviews.

What difficulty level is this interview question?

This is a medium difficulty Coding & Algorithms question, commonly asked during Technical Screen rounds at Microsoft.

What role is this question designed for?

This question is commonly asked for Machine Learning Engineer candidates at Microsoft during technical interviews.

Implement K-means clustering from scratch | Microsoft Interview Question

Quick Overview

This question evaluates competency in clustering algorithms, numerical computation, and practical algorithm implementation, focusing on core K-means concepts like distance-based grouping and centroid estimation. It is commonly asked because it exposes practical implementation ability, convergence reasoning, numerical edge-case handling (e.g.

Task: Implement K-means clustering (from scratch)

Write a function to perform K-means clustering on a set of points.

Input

A dataset X with shape (n_samples, n_features) (e.g., a NumPy array or PyTorch tensor).
An integer k = number of clusters.
Optional parameters:
- max_iters (e.g., 100)
- tol for convergence (e.g., 1e-4)
- Initialization method (e.g., random points from X )

Output

Return:

centroids : shape (k, n_features)
labels : length n_samples , where each label is in [0, k-1]

Requirements / Discussion Points

Describe and implement the two core steps iteratively:
1. Assignment step : assign each point to the nearest centroid (e.g., Euclidean distance).
2. Update step : recompute each centroid as the mean of assigned points.
Define a stopping condition (e.g., centroid shift < tol or max_iters reached).
Handle edge cases (at least verbally), e.g.:
- A cluster gets no points in an iteration (empty cluster).
- k > n_samples .
You may use Python with NumPy and/or PyTorch , but do not call a library K-means implementation.

(In the interview, correctness and reasoning matter more than making it fully runnable.)

Quick Overview

Task: Implement K-means clustering (from scratch)

Write a function to perform K-means clustering on a set of points.

Input

A dataset X with shape (n_samples, n_features) (e.g., a NumPy array or PyTorch tensor).

An integer k = number of clusters.

Optional parameters:

max_iters (e.g., 100)
tol for convergence (e.g., 1e-4)
Initialization method (e.g., random points from X )

Output

Return:

centroids : shape (k, n_features)

labels : length n_samples , where each label is in [0, k-1]

Requirements / Discussion Points

Describe and implement the two core steps iteratively:

Assignment step : assign each point to the nearest centroid (e.g., Euclidean distance).
Update step : recompute each centroid as the mean of assigned points.

Define a stopping condition (e.g., centroid shift < tol or max_iters reached).

Handle edge cases (at least verbally), e.g.:

A cluster gets no points in an iteration (empty cluster).
k > n_samples .

You may use Python with NumPy and/or PyTorch , but do not call a library K-means implementation.

(In the interview, correctness and reasoning matter more than making it fully runnable.)

Implement K-means clustering from scratch

Quick Overview

Task: Implement K-means clustering (from scratch)

Input

Output

Requirements / Discussion Points

Comments (0)

Implement K-means clustering from scratch

Quick Overview

Task: Implement K-means clustering (from scratch)

Input

Output

Requirements / Discussion Points

Comments (0)