- 30 min, 1 SQL problem, 1 Python problem, 1 reading comprehension/logic problem. All are super easy.
- 18 min, about 10 questions on machine learning.
- AIC vs. BIC, which one penalizes model complexity more
- Which cross validation technique to use for imbalanced data
- Recall vs. precision
- Shape of output for a single-class instance segmentation if the input tensor has shape (5,5,3)
- Which metrics does K-fold cross validation improves? (I know it improves bias and variance, but not sure about Type I or Type II error)
- Tree overfitting - 8 min, 4 questions on NLP
- TF vs. TFIDF
- What to do when SGD overfits (apply batch normalization? increase batch size?)
- Which optimization techniques gaurantee global maxima? (SGD, genetic algorithms, simulated annealing) - 8 min, 3 probability questions
- 60 people are divided into 6 groups, what is the probability that two friends are in the same group?
- what's the probability of 4 lyfts arriving before 3 uber given the distribution of their arrival times are the same?
Comments (1)