•               This plot shows the performance of learning algorithm for 1 serve over 10,000 trials
•                Here weights have been dumped and loaded in the successive run to tune the net
•                The number of iterations here are 3
•                The blue and pink lines are the final curves in the 3rd iteration.
•                The red and green lines are in the first iteration
•                Plota  where
•   a=1 refers to successful shots => reaching the ball
•   a=2 refers to completely successful shots => hitting a correct shot as well
•  Here
•    Gamma = 0.9
•     Lambda = 0.95
•     Reward Function  = 2  for complete success
                                                    = 1  for partial success
                                                    = -1 for failure
                                                    = 0 otherwise