Expected meetings in Room 1 after random assignment

Quick Overview

Evaluates conditional expectation in a balls-in-bins meeting-room assignment problem. Strong answers model Room 1 occupancy as binomial, compute non-empty probability, and derive expected meetings given Room 1 is non-empty.

Expected meetings in Room 1 after random assignment

Company: Meta

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Onsite

Scenario: Meeting rooms are randomly assigned. Given Room 1 already has at least one meeting, calculate its expected total when k meetings are spread across N rooms. ​ Question 1: Given Room 1 is non‑empty and k meetings are randomly assigned across N rooms, find expected meetings in Room 1. (Hint: conditional probability, linearity of expectation)

Quick Answer: Evaluates conditional expectation in a balls-in-bins meeting-room assignment problem. Strong answers model Room 1 occupancy as binomial, compute non-empty probability, and derive expected meetings given Room 1 is non-empty.

|Home/Analytics & Experimentation/Meta
Meta logo
Meta
Jul 12, 2025, 6:59 PM
mediumData ScientistOnsiteAnalytics & Experimentation
12
0

Expected Meetings in Room 1 Conditional on Being Non-empty

There are N rooms and k meetings. Each meeting independently chooses a room uniformly at random, and multiple meetings can be assigned to the same room. Let X1 be the number of meetings assigned to Room 1.

Given that Room 1 is non-empty, compute E[X1 | X1 >= 1].

Constraints & Assumptions

  • Each meeting assignment is independent.
  • Each room is equally likely for each meeting.
  • N and k are positive integers.
  • Use conditional expectation and linearity of expectation.

Clarifying Questions to Ask Guidance

  • Can multiple meetings be assigned to the same room?
  • Is Room 1 fixed in advance?
  • Are all rooms equally likely?
  • Are we conditioning on Room 1 specifically being non-empty or some room being non-empty?

What a Strong Answer Covers Guidance

  • Models X1 as Binomial(k, 1/N).
  • Computes E[X1] = k/N.
  • Computes P(X1 >= 1) = 1 - (1 - 1/N)^k.
  • Uses E[X | X > 0] = E[X] / P(X > 0) for nonnegative X.
  • Gives the exact expression (k/N) / [1 - (1 - 1/N)^k].
  • Checks edge cases such as k = 1 or large k.

Follow-up Questions Guidance

  • What is the expected number of non-empty rooms?
  • How would the result change if meetings choose rooms with unequal probabilities?
  • What is the probability Room 1 has exactly m meetings?
Loading comments...