Shulou(Shulou.com)06/03 Report--

Editor to share with you an example analysis of the basic concepts in the neural network, I believe that most people do not know much about it, so share this article for your reference, I hope you will learn a lot after reading this article. Let's learn about it!

Artificial neural network requires a certain mathematical basis, but generally speaking relatively simple, simple high mathematical foundation can, here sort out some of the most basic concepts needed to understand, for the introduction of neural network, very basic and important, and after understanding, you will find that the introduction does not need to look, sharpening the knife does not miss the firewood, it is strongly recommended to understand clearly after using sharp weapons such as tensorflow.

Independent variable / dependent variable / function

Because it is inevitable to come into contact with these contents when reading E documents, English is generally listed, and try to keep in mind that reading will greatly improve the speed.

Derivative

As the most basic derivative concept of high numbers, we will not repeat it here. We can simply understand the content and use a figure to explain it:

Basic concept

Derivative / partial derivative / directional derivative / gradient, these four concepts are extremely important, incomparably understood, simply sorted out as follows, for example, when it is difficult to understand the BP algorithm, please take this four concepts as the center to re-learn the relevant parts of the high number content.

Why the linear classification model can't deal with XOR problems

XOR is a very simple operation in a computer, while linear models such as perceptrons cannot solve the classification of XOR problems. For strict proof, please refer to Minsky's article.

In a nutshell, the classification according to the results can be divided into two categories (0ax 1), but if displayed on a plane, you will find that you can not find a straight line to directly separate the two types of results, so you can intuitively see that the linear classification model can not solve even the simple classification of XOR.

This sad view was proved in detail by Minsky in the famous Perceptron as early as 1969, which led to the freeze of artificial intelligence for about 10 years. This problem is not unsolved and can be solved by using two-layer perceptrons, but Minsky believes that it will lead to a huge amount of computation and there is no effective learning algorithm until the emergence of algorithms such as BP. In 1986, Learning representations by back-propagating errors published by Rumelhart,Geoffrey Hinton and Ronald Williams introduced the BP algorithm to the neural network model. By adding a hidden layer to the neural network and backpropagating the error at the same time, the amount of operation of error correction was reduced to a degree directly proportional to the number of neurons, which solved the problem of XOR classification of perceptrons and provided a learning algorithm for the model of multi-layer perceptrons.

Regression analysis.

In the introduction of the algorithm, we will learn to use linear regression and logical regression, so we need to have some understanding of regression analysis. Regression analysis studies the relationship between dependent variables and independent variables, which is widely used in prediction models. The number of independent variables / the type of dependent variables / the shape of the regression line all need to be considered. the common ways of regression analysis are as follows:

Linear Regression: linear regression

Logistic Regression: logical regression

Polynomial Regression: polynomial regression

Lasso Regression: lasso regression

ElasticNet Regression:ElasticNet regression

The above is all the contents of the article "sample Analysis of basic Concepts in Neural Networks". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.