Wednesday, 24 April 2019

np.random.randn() . vs. np.random.rand() vs. np.random.randint()

https://blog.csdn.net/zenghaitao0128/article/details/78556535

Below values are uniform distribution.  (what is uniform distribution:  https://www.zhihu.com/question/56891433)

This can be used to slow down gradient descent.


randn() .  will give you random value  from  -2.58 to +2.58

rand() will give you random value [0,1)

randint will return you value with given range

No comments:

Post a comment