Layered sampling tree method and device for fitting variable joint distribution
A joint distribution and variable technology, applied in the field of machine learning, can solve the problems of large differences in the overall distribution of the target sample set and cannot meet the scene simulation, and achieve the effect of improving the simulation accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
no. 1 example
[0047] see figure 1 .
[0048] Such as figure 1 As shown, this embodiment provides a hierarchical sampling tree method for fitting the joint distribution of variables, which at least includes the following steps:
[0049] S1. Obtain all characteristic variables with values of 0-1 in the sample data set, and arrange the characteristic variables according to the preset numbering sequence, and create a corresponding initial node structure;
[0050] S2. Traverse each sample in the sample data set, check the value of each feature of the sample according to the order of the characteristic variables, until all samples in the sample data set are checked, and generate a corresponding initial hierarchical sampling tree;
[0051] S3. Perform node correction on the initial hierarchical sampling tree until all nodes in the initial hierarchical sampling tree are traversed to obtain a corrected hierarchical sampling tree;
[0052] S4. A corresponding sample is generated each time throug...
no. 2 example
[0084] see figure 2 .
[0085] Such as figure 2 As shown, this embodiment provides a hierarchical sampling tree system for fitting the joint distribution of variables, including:
[0086] The initial node module 100 is used to obtain all characteristic variables with a value of 0-1 in the sample data set, and arrange the characteristic variables according to the preset numbering order to create a corresponding initial node structure;
[0087] For the initial node module 100, by obtaining all the characteristic variables with a value of 0-1 in the sample data set, assuming that the sample data has k characteristic variables with a value of 0-1, we arrange these variables in a certain number order, such as v 1 v 2 …v k , so as to create the corresponding initial node structure. Each layer of the sampling tree is a node containing the above k domains, and each domain contains a child node pointer and a count field (to be converted into a proportion in the end).
[0088] Th...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More