I am working on the MNIST dataset. How I can reduce 50% of this data?
(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()
Asked
Active
Viewed 256 times
1
Ethan
- 1,625
- 8
- 23
- 39
Salman Al-haddad
- 11
- 2
-
1What do you mean? Compress each image into $\dfrac{28\times 28}{2}=392$ features/pixels? Use $30000$ training images rather than the usual $60000?$ Use $5000$ testing images rather than the usual $10000?$ Some combination of these? Something totally different? – Dave Apr 13 '22 at 22:46
-
I want to use the total dataset as 35000 – Salman Al-haddad Apr 15 '22 at 20:08
-
What prevents you from just taking $35000$ images? // That’s still a decent number of images, and your training on that still won’t be immediate like a small OLS regression. Why shrink down to $35000?$ – Dave Apr 15 '22 at 21:10