ML
ML Interview Roadmap
← Study plan·Deep Learning

Day 40 of 133

DL loss functions + DSA Trees

Cross-entropy, label smoothing, focal, contrastive (InfoNCE), CTC.

DSA · NeetCode Trees

  • Validate Binary Search TreeDSA · Trees

    Interview questions to prep

    1. Why is checking 'left.val < node.val < right.val' insufficient — what's the right invariant?
    2. Compare in-order traversal vs min/max-bound recursion.
  • Kth Smallest Element IN A BstDSA · Trees

    Interview questions to prep

    1. Compare BFS vs DFS for this problem — which fits, and what's the iterative version?
    2. What's the recursion's space cost on the stack, and how would you go iterative if you needed O(log n)?
    3. What's the relationship between this problem's invariant and the BST property (if any)?

DL · Loss functions in deep learning

  • Cross-entropy, NLL, label smoothingDeep LearningGombru

    Interview questions to prep

    1. Why use cross-entropy instead of MSE for classification?
    2. What is label smoothing and why does it help calibration?
  • Focal, contrastive, triplet, InfoNCEDeep LearningLin et al.

    Interview questions to prep

    1. When does focal loss help, and how is it different from class weighting?
    2. What is contrastive loss (InfoNCE) and where is it used?
  • Sequence losses: CTC, seq2seq cross-entropyDeep LearningDistill

    Interview questions to prep

    1. Walk me through CTC loss for speech recognition.
    2. Why does CTC need a blank token, and how does the alignment marginalization work?

References & further reading