Model selection apparatus and method for data prediction

Abstract

Provided are a model selection device and method for data prediction. The model selection device includes: a matrix decomposition unit configured to perform singular value decomposition on the prediction matrix, wherein the row / column vectors of the prediction matrix are sample vectors, the column / row vectors are eigenvectors of corresponding dimensions, and one of the sample vectors is Containing the vector to be predicted of the feature sequence to be predicted; the model selection unit is configured to select the sub-matrix to be applied to the feature sequence to be predicted based on the sub-matrix of the sample vector that can reflect the sample vector obtained by the matrix decomposition unit by performing singular value decomposition predictive model. According to the solution disclosed in the present disclosure, the calculation dimension of the matrix can be reduced and the noise can be reduced, so as to achieve the effect of automatically and accurately selecting the prediction model.

Description

technical field [0001] The present disclosure generally relates to the field of data prediction, and in particular, relates to a model selection device and method for data prediction. Background technique [0002] In the field of data mining, for tasks such as prediction, classification, etc., there are many models available. For prediction tasks, for example, commonly available models include linear regression models, support vector regression models, and neural network-based models such as extreme learning machines. [0003] These models have their own applicable characteristics, and the effects are different for different data prediction tasks, even on different data sets. In general, certain models perform better than others on certain tasks and datasets. [0004] In order to achieve the best data prediction effect, it is often necessary to select an optional model in the early stage in order to use the best model that is most suitable for the current data set under th...

AI Technical Summary

Problems solved by technology

No matter which method is used, it depends on a certain degree of manual intervention, and often can only complete model selection through a small proportion of subset results on the complete data set, which has great limitations.

At the same time, the operation is cumbersome and time-consuming, but the effect may not be accurate enough

Images