Binary vs Multiclass Classification — ML Technical Interview Questions & Answers

Two-class problems vs three-or-more class problems.

Binary vs Multiclass Classification

  • Simple meaning: Binary = two classes (spam / not). Multiclass = three or more (red / yellow / green). Multilabel is different again — one email can be both “work” and “urgent”.
  • Picture: Pass/fail is binary. Grade A/B/C/D/F is multiclass. Don’t use a metric recipe from binary without checking (macro vs micro F1).
  • Some models are naturally multiclass (trees, softmax nets). Some are binary-first (classic SVM) and use one-vs-rest.
  • Trap: treating grades as a number 1–5 with linear regression and calling it classification. Trap: one-hot y by mistake in a model that wants 0,1,2.

Technical Interview Tip: Two classes vs many. Multilabel is ‘several tags at once’.

Code Examples for Binary vs Multiclass Classification

Binary vs Multiclass Classification — working example

def majority_vote(labels):
    return max(set(labels), key=labels.count)

neighbors = [1, 1, 0, 1, 0]
print(majority_vote(neighbors))  # kNN-style vote

# Logistic idea: score → probability via sigmoid
def sigmoid(z):
    return 1 / (1 + 2.718281828 ** (-z))
print(round(sigmoid(0.0), 3), round(sigmoid(2.0), 3))

Binary vs Multiclass Classification — classifiers map features to labels; kNN votes neighbors; logistic uses a sigmoid probability.

How to Explain Binary vs Multiclass Classification to an Interviewer

Binary example?

Spam vs ham Pass vs fail

Multiclass example?

Grades A–F Digit 0–9