classification Problem #581
-
|
I have just completed part 1 of Deep learning with tensorflow and I got a question that you have said earlier that we have to convert out data into tensors just like we did in our regression problem. But in classification problem, we didn't convert our data into tensors just used numpy array to train or test our model. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
|
Hey @AyeshaMalikAyesha , Good point! Yes, you're right, generally we should convert data to tensors but due to the flexibility of TensorFlow, it's often capable to compute on NumPy arrays as well. However, I'd recommend wherever possible using |
Beta Was this translation helpful? Give feedback.
Hey @AyeshaMalikAyesha ,
Good point!
Yes, you're right, generally we should convert data to tensors but due to the flexibility of TensorFlow, it's often capable to compute on NumPy arrays as well.
However, I'd recommend wherever possible using
tf.Tensoras this datatype will bring with it several advantages over NumPy arrays, see tensorflow.org/api_docs/python/tf/Tensor for more.