Which data type is explicitly required for independent variables in neural network modeling?

• A. Categorical
• B. Text
• C. Binary
• D. Numeric

Answer: (D)

Neural networks operate on numbers. The computations inside a network—multiplying inputs by weights, summing them, applying activation functions, and backpropagating errors—rely on numeric values and differentiable math. Because of that, the independent variables you feed into a network must be numeric data types.

If you start with non-numeric data, you convert it into numbers first. Categorical features are encoded (for example, one-hot or label encoding) so they become numeric vectors. Text data is transformed into numeric representations through methods like embeddings or vectorization. Binary data is already numeric (0 or 1), so it fits naturally. In short, the explicit requirement is numeric input, with non-numeric data needing conversion before modeling.