arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2205.13166stat.MLcs.ITcs.LGmath.IT
收藏

非现实环境下混合线性回归的学习

On Learning Mixture of Linear Regressions in the Non-Realizable Setting

Halıcıoğlu Data Science Institute (HDSI), UC San Diego(加州大学圣地亚哥分校哈利奇奥卢数据科学研究所);Google Research, India(谷歌研究院(印度));Google Research, USA(谷歌研究院(美国))

Avishek Ghosh, Arya Mazumdar, Soumyabrata Pal, Rajat Sen

摘要:

虽然混合线性回归(MLR)是一个很好的研究课题,以前的作品通常不分析这样的模型的预测误差。事实上,{ \em预测}和{ \em损失}在混合物的上下文中没有很好地定义。在本文中,我们首先展示了MLR可以用于预测,其中模型预测的不是标签,而是值列表(也称为{ \em列表解码})。列表大小等于混合物中组分的数目,损失函数被定义为所有组分模型导致的损失中的最小值。我们表明,与此定义,解决方案的经验风险最小化(ERM)实现小概率的预测误差。这需要一种算法来最小化MLR的经验风险,这在计算上是困难的。MLR中的现有算法工作集中于{ \em可实现}设置,即,当数据由混合线性(噪声)模型概率生成时,参数的恢复。在本文中,我们表明,一个版本的流行的交替最小化(AM)算法发现最佳拟合线的数据集,即使当一个可实现的模型是不假设的,在一些正则条件下的数据集和初始点,从而提供了一个解决方案的ERM。我们还提供了一个算法,在多项式时间内运行的数据点的数量,并恢复一个很好的近似的最佳拟合线。对这两种算法进行了实验比较。

英文摘要:

While mixture of linear regressions (MLR) is a well-studied topic, prior works usually do not analyze such models for prediction error. In fact, {\em prediction} and {\em loss} are not well-defined in the context of mixtures. In this paper, first we show that MLR can be used for prediction where instead of predicting a label, the model predicts a list of values (also known as {\em list-decoding}). The list size is equal to the number of components in the mixture, and the loss function is defined to be minimum among the losses resulted by all the component models. We show that with this definition, a solution of the empirical risk minimization (ERM) achieves small probability of prediction error. This begs for an algorithm to minimize the empirical risk for MLR, which is known to be computationally hard. Prior algorithmic works in MLR focus on the {\em realizable} setting, i.e., recovery of parameters when data is probabilistically generated by a mixed linear (noisy) model. In this paper we show that a version of the popular alternating minimization (AM) algorithm finds the best fit lines in a dataset even when a realizable model is not assumed, under some regularity conditions on the dataset and the initial points, and thereby provides a solution for the ERM. We further provide an algorithm that runs in polynomial time in the number of datapoints, and recovers a good approximation of the best fit lines. The two algorithms are experimentally compared.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑