Answer Happy

QUESTION NO. 02: You will do Forward propagation of Neural Networks using Multiprocessing and multithreading. Explanation of Forward Propagation of Neural Networks It consists of 3 steps, Step 1: multiply the weight matrix W with the features matrix Xi. Step 2: add the result of step 1 with biases matric b. the result will be called Z. Step 3: Apply the sigmoid function on all elements of the resultant matric from step 2. Sigmoid function is A=1/(1+e") 0.2 -0,5 0.1 2.0 56 1.1 -96.8 1.5 1.32.1 0.0 231 + 3.2 437.9 1/(1+e+(-96,8))=0 1/(1+e (4379) = 1 1/(1+e (6195)= 1 0 0 0.25 0.2 0.3 24 -1.2 61.95 W N b z A So, create a process P1, P2, and P3, pl will create multiple threads to multiply the weight matrix and biases matrix. (Matrix Multiplication using multithreading). The number of Threads in pl will be equal to the number of rows of weights W. P2 waits for its completion and uses the resultant matrix to add its biases matrix using multithreading (Matrix Addition using Multithreading). The number of Threads in p2 will be equal to the number of rows of biases matrix b. P3 weights for its completion and apply the sigmoid function on all elements of the resultant matrix from p2. P3 will also use multithreading to apply the sigmoid function. The number of Threads in p3 will be equal to the number of rows, resulting from p2. All threads work simultaneously to calculate the sigmoid function.