We built our own self-driving car. This is going to be a modelled version of a car (so it won't be driving on the streets of real cities) but still - it will learn how to drive itself. The key word here is learn, because the car will not be given any rules on how to operate in the environment before hand - it will have to figure everything out on its own. To achieve this, we will be using Deep Q-Learning. Deep Q-Learning is the result of combining Q-Learning with an Artificial Neural Network. The states of the environment are encoded by a vector which is passed as input into the Neural Network. Then the Neural Network will try to predict which action should be played, by returning as outputs a Q-value for each of the possible actions. Eventually, the best action to play is chosen by either taking the one that has the highest Q-value, or by overlaying a Softmax function.
