You are building a regression model for estimating the number of calls during an event. You need to determine whether the feature values achieve the conditions to build a Poisson regression model…

Question

Options

• AThe label data must be a negative value.
• BThe label data must be whole numbers.
• CThe label data must be non-discrete.
• DThe label data must be a positive value.
• EThe label data can be positive or negative.

• A
  8% (2)
• B
  88% (22)
• E
  4% (1)

Explanation

Poisson regression is designed for modeling count data-for example, the number of calls during an event. Its two core requirements for the label (target) are: (B) the values must be whole numbers (integers), since you cannot have a fractional count of events; and (D) the values must be non-negative (positive or zero), since counts cannot be negative. Negative values (A), non-discrete values (C), and unrestricted sign (E) all violate the assumptions of the Poisson distribution.

No community discussion yet for this question.