人工知能学会論文誌
Online ISSN : 1346-8030
Print ISSN : 1346-0714
ISSN-L : 1346-0714
原著論文
因子化漸近ベイズ推論による区分疎線形判別
藤巻 遼平山口 勇太郎江藤 力
著者情報
ジャーナル フリー

2016 年 31 巻 6 号 p. AI30-I_1-9

詳細
抄録

Piecewise sparse linear regression models using factorized asymptotic Bayesian inference (a.k.a. FAB/HME) have recently been employed in practical applications in many industries as a core algorithm of the Heterogeneous Mixture Learning technology. Such applications include sales forecasting in retail stores, energy demand prediction of buildings for smart city, parts demand prediction to optimize inventory, and so on. This paper extends FAB/HME for classification and conducts the following two essential improvements. First, we derive a refined version of factorized information criterion which offers a better approximation of Bayesian marginal log-likelihood. Second, we introduce an analytic quadratic lower bounding technique in an EM-like iterative optimization process of FAB/HME, which drastically reduces computational cost. Experimental results show that advantages of our piecewise sparse linear classification over state-of-the-art piecewise linear models.

著者関連情報
© 人工知能学会 2016
前の記事 次の記事
feedback
Top