Linear approximation abnormal lower limit determination method

Abstract

The invention relates to a linear approximation anomaly lower limit determination method, which is based on a rule that a test data background part obeys normal distribution and linear distribution must obey high normal distribution. The method comprises the following steps of writing an algorithm model to accurately separate a background part in the test digital set by using a statistical linearfitting successive approximation method through repetition (iteration) by means of a computer; wherein the upper limit and the lower limit of the separated background part are respectively the lower limit value of the positive anomaly and the upper limit value of the negative anomaly of the digital set, and the anomaly lower limit can be quickly, simply and accurately determined by the method.

Description

technical field [0001] The invention belongs to the field of data analysis research, and in particular relates to a method for determining the abnormal lower limit of "straight line approximation". Background technique [0002] Based on the fact that the background part of the test data obeys the law of normal distribution, there are nearly ten traditional methods for determining the lower limit of abnormality, and the calculation results of different methods vary greatly. So far, there is no method that convinces everyone, which seriously affects the research results. Uniformity and standardization. Inductively, there are two types of existing methods for determining the lower limit of anomalies: [0003] 1. Calculation method: "statistical iteration" calculation method ("The People's Republic of China Geology and Mineral Industry [0004] -Industry standard 1: 50,000 Geochemical Census Specifications (DZ / T 0011—91) "recommended method) as a representative, T = x ± kδ, (T...

AI Technical Summary

Problems solved by technology

[0009] 3. The traditional method of determining the lower limit of anomalies can only determine the lower limit of the positive anomaly of the number set, but cannot determine the upper limit of the negative anomaly

It does not conform to the three-point distribution characteristics of the test data set (positive anomaly, background, negative anomaly), such as the geological fact that the loss of geochemical elements cannot be studied, and the accurate research on data in the era of big data cannot be satisfied.

[0010] 4. The traditional method of determining the lower limit of abnormality, because the k value is uncertain (determined by the effect of the graph), the method of determining the lower limit of the graphic abnormality can only be drawn manually

Therefore, the traditional method of determining the lower limit of abnormalities cannot compile computer software, and cannot meet the processing requirements of large amounts of data with multiple maps and elements.

Images