相关实验视频

Updated: Jan 9, 2026

Cross-Modal Multivariate Pattern Analysis

Cross-Modal Multivariate Pattern Analysis

Published on: November 9, 2011

在类不平衡下使用CPRD数据评估XAI技术

Teena Rai¹, Jun He¹, Jaspreet Kaur²

¹Department of Computer Science, Nottingham Trent University, Nottingham, United Kingdom.

Frontiers in artificial intelligence

|December 1, 2025

概括

医疗保健数据中的阶级不平衡显著影响了LIME和SHAP等可解释的人工智能 (XAI) 方法的可靠性. 确保一致的模型解释对于临床决策支持系统中可靠的AI至关重要.

科学领域:

医疗保健人工智能的人工智能
机器学习的可解释性
临床决策支持临床决策支持

背景情况:

可解释的人工智能 (XAI) 对监管合规和对医疗保健的信任至关重要.
后期的XAI技术 (LIME,SHAP,PDPs) 被广泛用于解释医疗保健中的机器学习模型.
XAI技术的可靠性,特别是关于医学数据中的类不平衡,尚未完全理解.

研究的目的:

设计一个框架来评估阶级失衡对XAI解释的一致性的影响.
评估不同机器学习模型中的类失衡如何影响来自LIME,SHAP和PDP的解释.
在现实世界的临床数据场景中研究XAI技术的可靠性,其中有倾斜的类分布.

主要方法:

使用英国初级保健数据 (CPRD) 对LIME,SHAP和PDP的比较评估.
训练XGBoost,随机森林和MLP模型,在不平衡和平衡数据集上预测肺癌风险.
通过比较在不平衡数据和平衡数据上训练的模型来评估解释的一致性.

主要成果:

类失衡显著影响了LIME和SHAP解释的可靠性和一致性.
理论分析解释了为什么LIME和SHAP的解释会随着不同类分布而改变.
在阶级不平衡下,部分依赖图 (PDP) 也显示出临床相关特征的明显变化.

结论:

当前的XAI技术在应用于不平衡的医疗数据集时表现出脆弱性.
为了在医疗保健中可靠地部署机器学习,一致的模型解释至关重要.
解决类不平衡对于临床应用中可靠的XAI至关重要.

关键词:

对于CPRD来说,这是一个很大的问题.在 LIME 时代,在PDP中,PDP是PDP.这就是 SHAP SHAP 的意思.阶级不平衡阶级不平衡一致性的一致性不可解释的AI 评价评价评价评价

更多相关视频

An R-Based Landscape Validation of a Competing Risk Model

An R-Based Landscape Validation of a Competing Risk Model

Published on: September 16, 2022

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment

Published on: January 11, 2020

相关概念视频

Interpreting X̄ Charts

Interpreting X̄ Charts

Interpreting x̄ charts, a type of control chart used in statistical process control helps monitor the variation in processes over time. The x̄ chart is based on the sample mean and allows for monitoring variations in the process mean over time. These charts are pivotal for quality assurance in manufacturing and other sectors.
An x̄ chart plots the values of individual measurements over time against control limits calculated from historical data. The central line...

The X̄ Chart

The X̄ Chart

The x̄ chart is a statistical tool for monitoring the means in a process.
The x̄ chart, often known as the individual control chart, is a crucial tool in statistical process control. It is designed to monitor process behavior and performance over time and is widely used in various industries to ensure that processes are operating at their optimum capacity and within specified limits.
A x̄ chart is constructed by plotting individual measurements of a quality...

Testing a Claim about Population Proportion

Testing a Claim about Population Proportion

A complete procedure for testing a claim about a population proportion is provided here.
There are two methods of testing a claim about a population proportion: (1) Using the sample proportion from the data where a binomial distribution is approximated to the normal distribution and (2) Using the binomial probabilities calculated from the data.
The first method uses normal distribution as an approximation to the binomial distribution. The requirements are as follows: sample size is large...

The Ratio of X Chromosome to Autosomes

The Ratio of X Chromosome to Autosomes

In most organisms, sex is determined by the ratio of X and Y chromosomes. However, in some organisms, such as Drosophila and C.elegans, sex is determined by the ratio of the number of X chromosomes to the number of sets of autosomes. The Y chromosome in Drosophila is active but does not determine sex. It contains genes responsible for the production of sperms in adult flies.
Normal male Drosophila has a ratio of one X chromosome to two sets of autosomes. In contrast, normal female...

One-Way ANOVA: Unequal Sample Sizes

One-Way ANOVA: Unequal Sample Sizes

One-way ANOVA can be performed on three or more samples of unequal sizes. However, calculations get complicated when sample sizes are not always the same. So, while performing ANOVA with unequal samples size, the following equation is used:

Quantifying and Rejecting Outliers: The Grubbs Test

Quantifying and Rejecting Outliers: The Grubbs Test

Sometimes, a data set can have a recorded numerical observation that greatly deviates from the rest of the data. Assuming that the data is normally distributed, a statistical method called the Grubbs test can be used to determine whether the observation is truly an outlier. To perform a two-tailed Grubbs test, first, calculate the absolute difference between the outlier and the mean. Then, calculate the ratio between this difference and the standard deviation of the sample. This...