True or false: The neural network is both a predictive data mining model and a classification model.

• A. True
• B. False
• C. Not specified
• D. Sometimes

Answer: (A)

Neural networks are flexible function approximators used in data mining for both predicting numeric values and assigning instances to categories. They learn a mapping from input features to outputs by adjusting weights to minimize error on training data, which makes them a predictive model in general. For regression tasks, the network outputs a continuous value, often with a linear or identity activation at the final layer. For classification tasks, the final layer uses a probability-friendly activation like softmax (for multiple classes) or sigmoid (for binary), turning the network into a classifier. Because the same architecture can be trained for either objective by choosing the appropriate loss function and outputs, neural networks can serve as both predictive models and classification models.