Checkpoint1b TensorFlow2.txt - { nbformat :4 nbformat...

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":
{"name":"Checkpoint1b_TensorFlow2.ipynb","provenance":[],"collapsed_sections":
[]},"kernelspec":{"display_name":"Python
3","language":"python","name":"python3"}},"cells":
[{"cell_type":"markdown","metadata":{"id":"Fr39Cxns3cYv"},"source":["### COMP3359
Artificial Intelligence Applications\n","Department of Computer Science,
HKU\n","<br><br>\n","\n","# <u>Checkpoint 1b - TensorFlow 2</u> \n","\n","##
Estimated Time to Finish: \n","1~2 hours \n","\n","## Main Learning Objectives:
\n","-\tPractise usage of common ML framework to construct simple
application.\n","\n","## Overview \n","1.\t[Introduction](#s1) \n","2. [Before
You Start](#2)\n","3.\t[Preparation](#s3)\n","4.\t[Task - Reuse Model and Make
Predictions](#s2)\n","5.\t[Submission](#s5)\n","\n","-----\n"]},
{"cell_type":"markdown","metadata":{"id":"mlHp_N0OohlG"},"source":["<a
id=’s1’></a>\n","# 1 Introduction\n","\n","It may be a good idea to kick-start our
study of building AI applications by learning basic usage of some ML framework.
This checkpoint extends our other material \"ML Framework Learning Roadmap -
TensorFlow2\". The main task in this checkpoint is to reuse the model trained in
the example in \"ML Framework Learning Roadmap - TensorFlow2\", and give
predictions to data we provide. \n","\n","----- \n"]},
{"cell_type":"markdown","metadata":{"id":"9XOFPTzRtr7k"},"source":["<a
id=’s2’></a>\n","# 2 Before You Start\n","\n","## Referenced Tutorial\n","\n","This
checkpoint is mainly built by referencing the following tutorial. It is <b>strongly
recommended</b> to study it once first to understand the context of this
tutoiral.\n","\n","Referenced tutorial:\n","- [TensorFlow - Basic classification:
Classify images of clothing]
(
ials/keras/classification.ipynb#scrollTo=9ODch-OFCaW4)\n","\n","## Auxilliary
Tools\n","\n","In this checkpoint, you may need to use python packages to help you
tackle the problems. <b>If you have no experience</b> using the following packages,
it is <b>recommended</b> to check the following short tutorials and complete the
simple exercises inside.\n","\n","- NumPy\n"," - [NumPy UltraQuick Tutorial]
(-
edu/blob/master/ml/cc/exercises/numpy_ultraquick_tutorial.ipynb)\n","- Pandas\n","
- [Pandas DataFrame UltraQuick Tutorial]
(-
edu/blob/master/ml/cc/exercises/pandas_dataframe_ultraquick_tutorial.ipynb)\n","\n"
,"Since in the context of AI, we often handles a large number of numerical values
that are arraged as <b>(multi-dimensional) arrays</b> (e.g. vectors, matrices,
tensors), you may pay attentation to the manipulations of such (multi-dimensional)
arrays in these tutorials.\n","\n","------"]},{"cell_type":"markdown","metadata":
{"id":"meu7NxIRN1U5"},"source":["<a id=’s3’></a>\n","# 3 Preparation\n","\n","\"ML
Framework Learning Roadmap - TensorFlow2\" suggests a tutorial about training an