Regression prediction method and device

Abstract

The invention provides a regression prediction method, wherein not only similarity between independent variables X is taken into consideration, but also similarity between dependent variables Y of raw data is taken into consideration, and the model of output value y development based on the historical angle of close neighbors. Compared with the conventional model without taking data development mode into consideration, only one preprocessing section is added to a data set, and the information of the data point can be diversified without the need of extra resource; and the information of the raw data point X is diversified, and finally the prediction effect is improved. Furthermore, the regression prediction method can be realized on a MapReduce frame, and the execution speed can be improved by utilizing the parallelism of the device.

Description

technical field [0001] The invention belongs to statistical regression analysis and prediction, in particular to a regression prediction method and device used in statistical machine learning. Background technique [0002] Regression Analysis (Regression Analysis) is a method of statistically analyzing data, mainly to explore whether there is a specific relationship between data. Regression analysis is to establish a model of the relationship between the dependent variable Y (response variables) or dependent variables (dependent variables) and the independent variable X (predictors) or independent variables (independent variables). In statistical machine learning, the regression prediction method is mainly used to predict and analyze data. Among them, X is generally multidimensional data and Y is generally numerical data, which is called multiple regression. According to the regression equation, it can be divided into linear regression and nonlinear regression. The most b...

AI Technical Summary

Problems solved by technology

[0003] The existing regression prediction methods have the following two problems: First, due to the lack of data or feature selection, sometimes the original data points themselves may not contain enough information to perform regression prediction on the output (this problem can be referred to as feature missing); Secondly, because the data on each dimension of the data point X may not be numerical, it may not satisfy the changing law and range of numerical values, such as periodic angles, Boolean gender, etc., and enumerated colors Etc., which affects the effect of regression and the accuracy of prediction to a certain extent (this problem can be referred to as feature heterogeneity for short)

In addition, this method does not solve the problem of missing features and heterogeneous features

Images