Enterprise independent innovation ability prediction method based on support vector machine

Abstract

The invention provides an enterprise independent innovation ability prediction method based on a support vector machine. Firstly a hiberarchy model of dynamic parameter layer D is built, and relation between element group and index set of the dynamic parameter layer D is determined; the element group of the dynamic parameter layer D and the element group S of static parameter layer S form a sample set by certain pretreatment, and the sample set is divided into two parts of a training set and a testing set; four support vector machines of the model are respectively trained by the training set to obtain four regression functions, wherein the input is the element group of the dynamic parameter layer D and the output is one of the four elements in the element group S; the parameters therein are adjusted and optimized by the testing set to finalize the four regression functions; finally the model is applied to prediction of independent innovation ability of the sample in all directions. The model built by the invention has the advantages of high accuracy, fast processing speed and strong generalization ability, etc, and has a breakthrough in prediction.

Description

technical field [0001] The invention belongs to the field of artificial intelligence applications, and in particular relates to a method for predicting an enterprise's independent innovation capability based on a support vector machine. Background technique [0002] Since the reform and opening up, major changes have taken place in the level of technological development and the technical composition of the industrial structure in my country, and the quality of the labor force has also been considerably improved. But in general, in the face of the ever-changing scientific and technological changes, the increasingly intensified constraints on resources and the environment, and the fierce international competition characterized by innovation and technological upgrading, the problem of my country's weak independent innovation capabilities has increasingly become a problem of development. Bottleneck constraints. Accelerating the improvement of independent innovation capabilities ...

AI Technical Summary

Problems solved by technology

This method has two disadvantages: First, it needs to know the sample distribution form, which costs a lot of money, is sometimes difficult, and cannot be described in a low-dimensional space; second, the number of samples in traditional statistical research tends to infinity asymptotically theory, and most of the existing learning methods are based on this assumption

At present, there is no precedent for applying this theory to the prediction of independent innovation capability of enterprises.

Images