0 Basic Science TensorFlow (ii): First Knowledge Tensorflow_tensorflow

Objective

We successfully installed TensorFlow in the first article and introduced TensorFlow in our code. But for the 0 basis of me, still do not know what TensorFlow is (I believe a lot of people feel this level is very low), in this article describes how TensorFlow is coming, and what to do ~ TensorFlow is what

When you start to learn tensorflow, you will usually find the official website to study. For TensorFlow is what, the official website has a description:

Tensorflow™ is an open source software library that uses a data flow graph (stream graphs) for numerical computations. A node (Nodes) represents a mathematical operation in a graph, and a line (edges) in a graph represents an array of multidimensional data, the tensor (tensor), that is interconnected between nodes. Its flexible architecture allows you to expand computing on a variety of platforms, such as one or more CPUs (or GPU), servers, mobile devices, and so on in a desktop computer. TensorFlow was originally developed by researchers and engineers from the Google Brain group (Google Machine Intelligence Research) for machine learning and deep neural networks, but the versatility of the system makes it widely available in other computing areas.

After I understand the tensorflow and then read this paragraph, I feel very clear, but before use to see this, really did not understand what meaning. Then I'll explain it in the most popular words. TensorFlow, English literal translation, tensor flow. So what is tensor? As the A.zee book says:

A tensor is something which transforms like a tensor! A quantity that is transformed by a particular law under a different reference system is a tensor

For engineering people, it can be simply understood as a number of data, which may be matrices that may be arrays, and these data are transformed when using the TensorFlow tool.
So how do you understand flow? It can be imagined that the data is calculated every step.
For example, the TensorFlow can be understood as the production line of the automobile assembly line, which is the tool, our raw material is equivalent to the input, our car is equivalent to the output, and the car each step of the assembly, equivalent to the operation.
Overlooking the entire car shop, it's equivalent to seeing a data flow graph

Of course, TensorFlow also provides us with a lot of tools to help us get things done, such as motivating functions, normalization methods, gradient descent methods, which are helpful for machine learning. What TensorFlow can do?

He can do some analytical work instead of us, we can enter some data, through a series of calculations to draw some conclusions.
For medical workers, we can enter some images, such as CT, color ultrasound ... The images of cancer patients were analyzed and the findings of cancer risk were concluded.
For the financial Circle book, you can enter some previous stock data, training a model to analyze the future situation of stocks.
Of course, this intermediate algorithm is very important (this algorithm can be understood as the technical workers on the production line).
The above are individual analysis, may have incorrect place, welcome everyone to point out ~, just beginning contact, please forgive me.