Prediction model demonstration method and device and prediction model adjustment method and device

Abstract

The invention discloses a prediction model demonstration method and device and a prediction model adjustment method and device. The prediction model demonstration method comprises the following steps: obtaining at least one prediction result of a prediction model for at least one prediction sample; obtaining at least one decision-making tree training sample for training a decision-making tree model based on the at least one prediction sample and the at least one prediction result, wherein the decision-making tree model is used for fitting the prediction model; training the decision-making tree model by utilizing at least one decision-making tree training sample; and demonstrating the trained decision-making tree model visually. The method can enable the prediction model hard to understand to be approximated to the decision-making tree model, and demonstrate the approximated decision-making tree model, so that a user can better understand the prediction model based on the demonstrated decision-making tree model.

Description

technical field [0001] The present invention relates to the field of data science, in particular to a method and device for displaying a prediction model and a method and device for adjusting a prediction model. Background technique [0002] In traditional data mining methods, statistical methods are often used to conduct regression analysis or factor analysis on historical data, and then analyze the causes of historical data, and use the obtained rules to build a decision-making engine. In statistical learning theory, structured risk is equal to the sum of empirical risk and belief risk. Among them, experience risk is the error of learned knowledge on a given sample set, while confidence risk is the error of knowledge on unknown samples. Therefore, traditional statistical methods only minimize the risk on known samples, and cannot guarantee the suppression of confidence risks on unknown samples. That is to say, from the perspective of statistical learning, this method onl...

AI Technical Summary

Problems solved by technology

[0020] The above are just examples of deep learning neural network algorithms. In fact, even technicians in the field of machine learning have difficulty in intuitively understanding how the machine learning model answers questions from the mathematical expression of the machine learning model.

That is to say, compared with the traditional data mining model, which can use statistical methods to interpret and decompose the model, it is currently difficult to effectively explain and analyze the machine learning model, resulting in it being unable to be better used by users (for example, business experts, business executives, etc.) interpreted and understood

For example, because the machine learning model for disease diagnosis itself is difficult to understand, it will cause people to question the reliability of the diagnosis and it is difficult to promote

[0021] In reality, not only the pure machine learning model has the above defects, but with the integration of machine learning and statistical learning, many other types of models (such as expert systems and other models based on many rules or factors to answer questions) also unexplained problem

People's lack of understanding of the model will hinder the practical application of the model, for example, it is difficult for users to understand the model well, making it difficult to effectively use the model to solve business problems

Images