Given a dataset where the dataset is not linearly-separable, and each of the fea- tures have continuous values, which of the following algorithms is more ideally suited a) perceptron b) decision-trees c) neural-networks. Why

Option c neural-networks

Explanation:

Since the dataset is not linearly-separable, the perceptron is not a good option.  A single layer perceptron can only work on linearly separable dataset. On another hand, since the features have continuous values, the decision-trees is not a good option. The continuous values will result in a extremely complicated tree structures which may expect a very high computational cost for a simple prediction. The most ideal choice is neural-networks at it can learn the complicated non-linear pattern and produce the optimum predictive model.

this seems to be in the wrong category. but i'll take some free points my dude

posting this in wrong category will result in people seeing it who won't know the answer.