Skip to content
RED SAGSkills

Easy Machine Learning Engineer questions

34 questions · 5 points each for a correct first answer.

easymediumhard
  1. 1.Big gap between train and validation accuracyeasy
  2. 2.Why keep a separate test seteasy
  3. 3.Which model needs feature scalingeasy
  4. 4.Definition of precisioneasy
  5. 5.Symptom of a learning rate that is too higheasy
  6. 6.Encoding a city featureeasy
  7. 7.Bias and variance as complexity growseasy
  8. 8.Accuracy on imbalanced dataeasy
  9. 9.Shape of a matrix producteasy
  10. 10.Cosine similarity by handeasy
  11. 11.Chain rule on a squared termeasy
  12. 12.What sampling temperature doeseasy
  13. 13.Standard error of a sample meaneasy
  14. 14.What logistic regression outputseasy
  15. 15.k-nearest neighbours voteeasy
  16. 16.What k-means minimiseseasy
  17. 17.Variance kept by two componentseasy
  18. 18.Gini impurity of a nodeeasy
  19. 19.Why random forests average treeseasy
  20. 20.ReLU applied to a vectoreasy
  21. 21.Recall from confusion countseasy
  22. 22.RMSE from four errorseasy
  23. 23.Broadcasting a column and a roweasy
  24. 24.Dropout after eval()easy
  25. 25.Why export a model to ONNXeasy
  26. 26.Why LLMs use subword tokenseasy
  27. 27.What text embeddings captureeasy
  28. 28.Standardising a single valueeasy
  29. 29.When early stopping triggerseasy
  30. 30.Optimizer steps in one epocheasy
  31. 31.Expected starting loss, 10 classeseasy
  32. 32.What retrieval-augmented generation addseasy
  33. 33.What data drift meanseasy
  34. 34.What prompt injection iseasy