Design a decision tree

Nonlinear systems do not obey the proportionality relation between input and output changes as do linear systems. Consider an artificial neuron with a threshold ? = 0.50. Argue that this neuron is a nonlinear system. Why might stress in humans be considered a nonlinear phenomenon?

In Section 1 we portray an ANN as a black box. What limitations does this opaqueness impose on their utility? In a linear system, the output is proportionally related to the input, in other words, small changes in the input produce correspondingly small changes in output and similarly for large input changes.

Describe two systems in nature that are examples of linear systems.

A single-layer neural network cannot implement a function that is not linearly separable. Is this a serious drawback? Explain. 6. The learning rate is a constant between 0 and 1, in other words, 0

Why does choosing that attribute with the largest information gain favor construction of shorter decision trees? Suggest a possible method to handle attributes with continuous values.

Design a decision tree for the following Boolean functions: a. a ? (b ^ ~ c) b. majority (x,y,z) Calculate the entropy for each of the following sets: a. [6(+), 11(-)] b. [1(+), 9(-)] c. [2(+), 12(-)]

Solution: