Buddy and Experiments

Buddy is new to Machine Learning. So he starts with gradient descent. While learning, He came up with an idea to experiment with. He starts with an initial parameter A, an expected parameter B, and a learning rate C. At every step, he adds to A, C times the sum of A and B. More formally at every step he does A = A+ C(B+A). He realizes that this grows very fast and soon goes out of his control. S