Classifying a Stream of Infinite Concepts: A Bayesian Non-parametric Approach

Autor:	Hassan Hafez, Ali Soltani-Farani, Seyyed Abbas Hosseini, Hamid R. Rabiee
Rok vydání:	2014
Předmět:	Data stream Concept drift Computer science business.industry Data stream mining Bayesian probability Probabilistic logic Inference Machine learning computer.software_genre Generative model Data mining Artificial intelligence Online algorithm business computer
Zdroj:	Machine Learning and Knowledge Discovery in Databases ISBN: 9783662448472 ECML/PKDD (1)
DOI:	10.1007/978-3-662-44848-9_1
Popis:	Classifying streams of data, for instance financial transactions or emails, is an essential element in applications such as online advertising and spam or fraud detection. The data stream is often large or even unbounded; furthermore, the stream is in many instances non-stationary. Therefore, an adaptive approach is required that can manage concept drift in an online fashion. This paper presents a probabilistic non-parametric generative model for stream classification that can handle concept drift efficiently and adjust its complexity over time. Unlike recent methods, the proposed model handles concept drift by adapting data-concept association without unnecessary i.i.d. assumption among the data of a batch. This allows the model to efficiently classify data using fewer and simpler base classifiers. Moreover, an online algorithm for making inference on the proposed non-conjugate time-dependent non-parametric model is proposed. Extensive experimental results on several stream datasets demonstrate the effectiveness of the proposed model.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::4d230e25050480dfb9cce42af22f2627 https://doi.org/10.1007/978-3-662-44848-9_1 Zobrazit plný text záznamu