Shulou(Shulou.com)06/02 Report--

In this issue, the editor will bring you about the role of data normalization processing. The article is rich in content and analyzed and described from a professional point of view. I hope you can get something after reading this article.

The purpose of data normalization processing is to limit the preprocessed data to a certain range, so as to eliminate the adverse effects caused by singular sample data. After data normalization, the speed of gradient descent to find the optimal solution can be accelerated, and the accuracy may be improved (such as KNN).

The purpose of normalization is to limit the preprocessed data to a certain range (such as [0d1] or [- 1d1]), so as to eliminate the adverse effects caused by singular sample data.

The main results are as follows: 1) in statistics, the specific function of normalization is to induce the statistical distribution of unified samples. Normalized between 0: 1 is the probability distribution of statistics, and normalized between-1 is the coordinate distribution of statistics.

2) singular sample data refers to the sample vector (eigenvector) that is particularly large or small relative to other input samples, for example, the following sample data x1, x2, x3, x4, x5, x6 (eigenvector-> column vector), in which the two features of x6 are quite different from other samples, so x6 is considered to be singular sample data.

The existence of singular sample data will not only increase the training time, but also may lead to non-convergence, so when there is singular sample data, the preprocessed data need to be normalized before training; on the contrary, when there is no singular sample data, it can not be normalized.

If the normalization is not carried out, the objective function will become "flat" because the values of different features in the feature vector are quite different. In this way, when the gradient is descending, the direction of the gradient will deviate from the direction of the minimum and take many detours, that is, the training time is too long.

After normalization, the objective function will be relatively "round", so that the training speed will be greatly accelerated and there will be fewer detours.

To sum up, normalization has the following advantages, namely

The main results are as follows: 1) after normalization, the speed of gradient descent to find the optimal solution is accelerated.

2) normalization may improve accuracy (such as KNN)

Note: there is no data standardization method, put on every problem, put on each model, can improve the accuracy of the algorithm and accelerate the convergence speed of the algorithm.

The above is the role of the data normalization processing shared by the editor. If you happen to have similar doubts, you might as well refer to the above analysis to understand. If you want to know more about it, you are 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.