added gradient descent code

gradient_descent.py

@@ -0,0 +1,18 @@
+import numpy as np
+def gradient_descent(x,y):
+    w_curr=b_curr=0
+    iterations=10000
+    n=len(x)
+    learning_rate=0.16
+    for i in range(iterations):
+        y_predicted=w_curr*x+b_curr
+        cost=(1/2*n)*sum([val**2 for val in (y_predicted-y)])
+        dw=(1/n)*sum(x*(y_predicted-y))
+        db=(1/n)*sum(y_predicted-y)
+        w_curr-=learning_rate*dw
+        b_curr-=learning_rate*db
+        print(f'w:{w_curr} b:{b_curr} cost:{cost} iterations:{i+1}')
+
+x=np.array([1,2,3,4,5])
+y=np.array([5,7,9,11,13])
+gradient_descent(x,y)