Neural Networks and TensorFlow - Deep Learning Series [Part 8]

---

Building on top of the previous lesson, a typical project in TensorFlow is divided into two stages:

• creating the computational graph
• executing the graph.

In the first stage we define all the variables, placeholders, constant, etc and the operations between them.

In the second stage or phase we run the graph with all the definitions and operation within a TensorFlow session. Typically, when we create the session, the first thing we need to do is to initialize all the variables and only thereafter execute the graph.

In the lesson below we run through a simple example that includes all of these concepts. The data we're going to use is generated using the random function in numpy.

---

---

To stay in touch with me, follow @cristi

---

Cristi Vlad Self-Experimenter and Author