SchoolWhool
Quizzes and saved progress, freeSign in freeSign in
Machine Learning · Module 1Module material

What Machine Learning Is interview questions

15 questions interviewers ask on this module’s topics, from first principles to the follow-ups. Each comes with a model answer.

What learning is

  1. What is machine learning, in one sentence, and what is a parameter?●
  2. What is a baseline, and why report one?●
  3. When would you not use machine learning?●
  4. A model scores higher on the test set than on the training set. Is something wrong?●●

Kinds of problem

  1. How do you decide between regression and classification?●
  2. What is the difference between supervised and unsupervised learning? Give one task of each.●
  3. Can the same column be a regression target and a classification target?●●

The workflow

  1. List the steps of a machine learning project. Why does the split come before exploring?●
  2. A model scores 100% on its training data. What do you check?●●
  3. Why not choose between model versions on the test set?●●

Data

  1. What are X and y, and what shapes do they have for Auto MPG?●
  2. A column called origin holds 1, 2 and 3, and its mean is 1.57. What does that tell you?●
  3. Why remove duplicates before splitting, not after?●●
  4. What does label noise do to a model? Is all noise equally harmful?●●
  5. What is a collection artifact? Give an example.●●