Project Team : Ankit Kumar(121ad0005@iiitk.ac.in) & Rajiv Kumar(121ad0043@iiitk.ac.in)
Solving real life problem faster and correctly with the AI&ML models helps us in many ways . So that's why we are modifying Dueling DQN structure by splitting the dense layer (more than one) and create their own separate advantage and value streams that converge at the final layer , allowing parallel training: one for estimating the state value V(s) and one for estimating the advantage for each action A(s,a). This will increase the convergence rate means this can lead to faster training and quicker performance improvement .This will increase our correctness in problem solving and also in Robotics and gaming for taking action.
Back