IAE Academy

ISRO CS 2023 Q.62: solution and answer

ISRO 2023 · CSQ.62 · 1 mark · MCQ

A hybrid Bayesian network contains

  1. (A)

    Only continuous variables

  2. (B)

    Only discrete variables

  3. (C)

    Both discrete and discontinuous variables

  4. (D)

    Both continuous and discrete variables

See the question in the paper
ISRO CS 2023 question 62Tap to open full size

Answer and solution

ANSWER

A hybrid Bayesian network contains both continuous and discrete variables, which is option (d).

Official answer: DSource: ISRO's revised and final answer key of 16 February 2024

  1. A Bayesian network is a directed acyclic graph whose nodes are random variables, each with a conditional distribution given its parents.

  2. In a discrete network every node takes values from a finite set. In a continuous network every node is real-valued, for example linear-Gaussian.

  3. A hybrid Bayesian network mixes the two kinds. Examples are a continuous node with a discrete parent (a conditional Gaussian) and a discrete node with a continuous parent (a threshold or logistic model).

  4. So option (a), only continuous, and option (b), only discrete, describe the pure cases, not the hybrid case.

  5. Option (c) is wrong because 'discontinuous' is not a variable type here. The standard contrast is continuous versus discrete.

  6. Option (d) states the definition: both continuous and discrete variables.

KEY IDEA

"Hybrid" means the network mixes the two variable types: continuous nodes (often Gaussian) and discrete nodes, linked by conditional distributions such as conditional Gaussians or threshold models.

WATCH OUT FOR

Picking option (c) because it also says "both". The pair is continuous and discrete, not discrete and "discontinuous", which is not a variable type.

From the ISRO CS 2023 paper. Every question of this paper · The full paper (PDF, ISRO) · All 1 ISRO Uncertainty question, with notes