Would you please solve the Machine Learning problem with a proper answer?

3. Derive the perceptron training rule and gradient descent training rule for a single unit with output o, where o=w0+w1x1+w1x12++wnxn+wnxn2. What are the advantages of using gradient descent training rule for training neural networks over the perceptron training rule