Chapter 1: Modeling Procedure of TensorFlow

Although Tensorflow is designed in a smart way to be adaptive to various complex numerical computations, the most popular usage is implementation of machine learning models, especially for those of neural networks.

In principle, the neural network could be defined by graphs consist of tensors and trained through automatic differenciate.

However, for simplification, we recommend to use high-level Keras API in Tensorflow to implement the neural networks.

The common procedures of implementing neural networks using TensorFlow are:

  1. Data preparation

  2. Model definition

  3. Model training

  4. Model evaluation

  5. Model application

  6. Model saving

For the beginners, actually, data preparation is the most difficult part.

The most common data types are structured data, images, texts, and temporal sequences.

We are demonstrating the steps of modeling for these four data types through the following examples, respectively: (1) Predicting the survival on the Titanic; (2) Image classification on CIFAR2 set; (3) Classification of movie reviews on IMDB; (4) Predicting the terminate date of the COVID-19 pandemic in China.

Please leave comments in the WeChat official account “Python与算法之美” (Elegant Python and Algorithms) if you want to communicate with the author about the content. The author will try best to reply given the limited time available.

You are also welcomed to reply 加群(join group) in the WeChat official account to join the group chat with the other readers.

image.png