An Online EM Algorithm Using Component Reduction
نویسندگان
چکیده
The EM algorithm has been widely used in many learning or statistical tasks. However, since it requires multiple database scans, applying the EM algorithm to data streams is not straight forward. In this paper we propose an online EM algorithm which can deal with data streams. The algorithm utilizes a component reduction technique which reduces the number of components in a mixture model. A notable advantage of our algorithm over existing online variants of the EM algorithm lies in its simplicity. Our algorithm almost preserves the theoretical clearness of the EM algorithm.
منابع مشابه
Online Composition Prediction of a Debutanizer Column Using Artificial Neural Network
The current method for composition measurement of an industrial distillation column includes an offline method, which is slow, tedious and could lead to inaccurate results. Among advantages of using online composition designed are to overcome the long time delay introduced by laboratory sampling and provide better estimation, which is suitable for online monitoring purposes. This paper pres...
متن کاملOnline Torque Ripple Reduction in a Three-Phase 12 by 8 Switched Reluctance Motor Using Genetic Algorithm in PWM Generation
Despite a large number of advantages, Torque Ripple (TR) is the most important drawback of Switched Reluctance Motor (SRM). In the presented study, TR was reduced by optimizing the gate pulse angle of the SRM phase which played a leading role in the generated torque profile. For the Optimization, one of the strategies of Genetic Algorithm (GA) which was named Non-dominated Sorting Genetic Algor...
متن کاملIndependent Vector Analysis for Source Separation Using a Mixture of Gaussians Prior
Convolutive mixtures of signals, which are common in acoustic environments, can be difficult to separate into their component sources. Here we present a uniform probabilistic framework to separate convolutive mixtures of acoustic signals using independent vector analysis (IVA), which is based on a joint distribution for the frequency components originating from the same source and is capable of...
متن کاملApproximate Learning of Dynamic Models
Inference is a key component in learning probabilistic models from partially observable data. When learning temporal models, each of the many inference phases requires a traversal over an entire long data sequence; furthermore, the data structures manipulated are exponentially large, making this process computationally expensive. In [2], we describe an approximate inference algorithm for monito...
متن کاملRejection of the Feed-Flow Disturbances in a Multi-Component Distillation Column Using a Multiple Neural Network Model-Predictive Controller
This article deals with the issues associated with developing a new design methodology for the nonlinear model-predictive control (MPC) of a chemical plant. A combination of multiple neural networks is selected and used to model a nonlinear multi-input multi-output (MIMO) process with time delays. An optimization procedure for a neural MPC algorithm based on this model is then developed. T...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2004