|
||||||||
Handwritten Digit Recognition - Python Automation and Machine Learning for ICs - - An Online Book - |
||||||||
Python Automation and Machine Learning for ICs http://www.globalsino.com/ICs/ | ||||||||
Chapter/Index: Introduction | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | Appendix | ||||||||
================================================================================= TensorFlow provides various datasets through its tf.keras.datasets module. The MNIST dataset is a commonly used dataset for handwritten digit recognition (code): Input: Output when epochs=1:
This is a wrong prediction with just 1 epoch. Output when epochs=2:
This is still a wrong prediction with just 2epoch. Output when epochs=5:
Now, the prediction is correct with an increased numbers of epochs (5 epoch). The code (x_train, y_train), (x_test, y_test) = mnist.load_data() above is used to load the MNIST dataset in TensorFlow:
============================================
|
||||||||
================================================================================= | ||||||||
|
||||||||