An honest cram sheet for ML and computer vision interviews

What it does

This repository is a curated document of roughly 65 technical interview questions and model answers for machine-learning and computer-vision engineering positions. Each entry links to an external source—blogs, Wikipedia, or academic papers—so the explanations are aggregated rather than original. A recently added section also covers natural-language processing questions.

The interesting bit

The collection treats computer vision as a first-class topic rather than an afterthought, covering receptive-field arithmetic, sparse-matrix classes in C++, integral-image algorithms, and RANSAC outlier removal alongside standard deep-learning theory.

Key highlights

• Covers classic ML concepts like bias-variance tradeoffs, regularization, PCA, and gradient descent with concise explanations
• Includes computer-vision specifics such as CNN receptive fields, connected-component labeling, and content-based image retrieval
• Recently expanded to include NLP interview questions for 2026
• Every answer cites an external source, functioning more as a bibliography than a textbook
• Sponsored by a job-application automation tool, which reveals the intended audience

Caveats

• This is a curated list, not a framework or codebase; there are no runnable examples or notebooks
• Several explanations contain visible typos and rough phrasing that suggest minimal editing
• Some answers in the README are truncated, and the full depth often requires following links to external sites

Verdict

Grab it if you are cramming for an ML or CV interview and need a single page of talking points to cross-reference. Skip it if you want hands-on projects, original pedagogy, or executable code.