Problem Description You are tasked with developing a Q$-$learning agent to solve a grid world environment using reinforcement learning and Python. The grid world is represented as a $5$x$5$ grid, and the agent must navigate through it$,$ avoiding obstacles, and reach the terminal state to receive a reward. Grid World Configuration and Rules The grid world is a $5$x$5$ matrix bounded by borders. The agent starts from cell $[2,1]($second row, first column$).$ The agent has four possible actions: North $($action code: $-1)$ South $($action code: $-2)$ East $($action code: $-3)$ West $($action code: $-4)$ The agent receives a reward of $+10$ if it reaches the terminal state cell $[5,1]($blue cell$).$ There is a special jump from cell $[4,2]$ to cell $[4,4]$ with a reward of $+5.$ The agent is blocked by obstacles $($black cells$).$ Q$-$Learning Approach Q$-$learning is a model$-$free reinforcement learning algorithm that learns an action$-$value function $($Q$-$values$)$ for each state$-$action pair. Here$$s how you can approach this task: Initialization: Initialize the Q$-$values for all state$-$action pairs to arbitrary values $($e$.$g$.,$ zeros$).$ Set the learning rate $(\backslash$alpha $)$ and discount factor $(\backslash$gamma $).$ Exploration and Exploitation: Exploration: The agent explores different actions to discover the environment. Use an exploration strategy $($e$.$g$.,\backslash$epsi$-$greedy$)$ to choose actions randomly with some probability. Exploitation: The agent exploits the learned Q$-$values to choose the best action based on the current state. Q$-$Value Update: Update the Q$-$values using the Q$-$learning update rule:Q$($s$,$a$)$Q$($s$,$a$)+\backslash$alpha $($r$($s$,$a$)+\backslash$gamma a$$max$$Q$($s$,$a$)$Q$($s$,$a$))$ where: $($s$)$ is the current state. $($a$)$ is the chosen action. $($s$)$ is the next state after taking action $($a$).($r$($s$,$ a$))$ is the immediate reward for taking action $($a$)$ in state $($s$).(\backslash$alpha$)$ is the learning rate. $(\backslash$gamma$)$ is the discount factor. Training the Agent: Run episodes where the agent interacts with the environment. Update Q$-$values based on observed rewards and transitions. Continue until convergence or a maximum number of episodes. Policy Extraction: Extract the policy $($optimal action for each state$)$ from the learned Q$-$values. Use the policy to navigate the agent through the grid world. Remember to handle special cases $($e$.$g$.,$ the jump from cell $[4,2]$ to $[4,4])$ appropriately in your implementation.