Select Page

Keras for Machine Learning Engineers With 10 Simple Steps

TensorFlow, Numpy, Keras, Machine Learning for Engineers - BitColon.com for ML and Data Science Engineers, Scientists & Engineers - Owned & Managed by Factober

Written by bitcolon

Editor & Administrator bitcolon.com (Part of Factober Blogging Network)

September 6, 2020

Your Title Goes Here

Your content goes here. Edit or remove this text inline or in the module Content settings. You can also style every aspect of this content in the module Design settings and even apply custom CSS to this text in the module Advanced settings.

N

1. Setup Development Environment (System Requirements)

The tutorial is performed on:

  • Ubuntu 20.04
  • Python 3

Install Tensorflow [We will import Keras from Tensorflow]

# Requires the latest pip
pip install –upgrade pip

# Current stable release for CPU and GPU
pip install tensorflow

# Or try the preview build (unstable)
pip install tf-nightly

    N

    2. Vectorized and standardized data

    Neural Networks don’t process raw data. So, well prepared and organized data should be supplied to them in vectors and standardized representations.

    Text

    1. Read the text files into string tensors.
    2. Split those strings into words.
    3. Index the words.
    4. Turn the indexed words into integer tensors.

    Images

    1. Read the image files.
    2. Then, decoded them into integer tensors.
    3. Then, convert them into floating-points.
    4. Then, Normalize them into small values (usually between 0 and 1).

    CSV (Comma Separated Values)

    1. Parse the CSV data
    2. Convert the numerical features to floating point tensors.
    3. Index the categorical features.
    4. Then conver them to integer tensors.
    5. Then normalize each feature to zero-mean and unit-variance.
    N

    3. Loading the Data

    Keras models accept data in the following formats:

    NumPy Array

    1. Read the text files into string tensors.
    2. Split those strings into words.
    3. Index the words.
    4. Turn the indexed words into integer tensors.

    tf.data

    tf.data is is a tensorflow API that allows to build complex input pipelines from simple, reusable pieces. Read more about tf.data API here.

    Python Generators

    Python generators are the functions that return an object (iterator) which we can iterate over (one value at a time).

    Is it compulsory to learn JavaScript along with HTML and CSS?

    Before going any further let's discuss, what JavaScript? And, simple answer is, that, there is not other alternative to it. For layman, JavaScript is some kind of computer language that add interactivity to components or elements you see on web page.  These web pages...

    HTML code to display heading computer exam in largest size

    HTML a.k.a Hypertext Markup Language, is the standard scripting language for defining markups on web pages. Later on, you can style those markups with CSS or add life to them with JavaScript. HTML has its own predefined semantics for common elements, like, paragraphs,...

    Difference between HTML document and HTML program

    Before, getting into the topic, make sure you clear you what is HTML? HTML is Hypertext Markup Language. It falls into the category of scripting languages. Programs written using these languages are not compiled, but directly feed to some kind of interpreter, and it...

    You May Also Like…

    No Results Found

    The page you requested could not be found. Try refining your search, or use the navigation above to locate the post.

    0 Comments

    Submit a Comment

    Your email address will not be published. Required fields are marked *

    Pin It on Pinterest

    Share This