Bias/Variance Analysis for Relational Domains
Bias/Variance Analysis for Relational Domains
复制标题
DOI：
10.1007/978-3-540-78469-2_6
复制
发表时间：
2007-06
期刊：
影响因子：
--
通讯作者：
Jennifer Neville;David D. Jensen
中科院分区：
文献类型：
--
作者：
Jennifer Neville;David D. Jensen
偏差/方差分析[1]是研究机器学习算法性能的有用工具。传统的分析将损失分解为由于学习过程的各个方面而引起的误差，其基本假设是，由于用于预测的推理过程，模型预测没有变化。在关系数据中，当相关实例的类标签之间存在依赖关系时，对一个对象的推断可以用来改进对其他相关对象的推断。集体推理技术通过联合推理测试集中的类标签来利用这些依赖关系。这种方法可以为每个实例独立地产生比条件推理更准确的预测，但它也引入了一个额外的错误源，通过使用近似推理算法和通过测试集信息的可用性变化。迄今为止，尚未研究引用错误对关系模型性能的影响。
Bias/variance analysis [1] is a useful tool for investigating the performance of machine learning algorithms. Conventional analysis decomposes loss into errors due to aspects of the learning process with an underlying assumption that there is no variation in model predictions due to the inference process used for prediction. This assumption is often violated whencollective inferencemodels are used for classification of relational data.In relational data, when there are dependencies among the class labels of related instances, the inferences about one object can be used to improve the inferences about other related objects. Collective inference techniques exploit these dependencies byjointlyinferring the class labels in a test set. This approach can produce more accurate predictions than conditional inference for each instance independently, but it also introduces an additional source of error, both through the use of approximate inference algorithms and through variation in the availability of test set information. To date, the impact ofinferenceerror on relational model performance has not been investigated.