Logistic Regression — ML Technical Interview Questions & Answers

A linear model for binary classification problems.

Logistic Regression

  • Simple meaning: Despite the name, logistic regression is classification. It outputs a probability of “pass” (or spam), then you pick a threshold (often 0.5) to say yes/no.
  • Picture: hours studied → P(pass). Sigmoid squashes any number into 0..1. Asha 2 hours → 0.18 fail. Kabir 6 hours → 0.91 pass.
  • sklearn: LogisticRegression().fit then predict or predict_proba. On imbalanced pass/fail, don’t brag accuracy — use precision/recall/F1.
  • Trap: “logistic is regression because of the name”. Trap: 95% pass dummy accuracy.

Technical Interview Tip: Name says regression, job is classification. Probability + threshold.

Code Examples for Logistic Regression

Logistic Regression — 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))

Logistic Regression — classifiers map features to labels; kNN votes neighbors; logistic uses a sigmoid probability.

How to Explain Logistic Regression to an Interviewer

Is logistic regression for numbers or classes?

Classes Probability of a label

What is the sigmoid for?

Squash to a probability 0 to 1