A (stochastic) Em in General

نویسندگان

  • Manzil Zaheer
  • Michael Wick
  • Jean-Baptiste Tristan
  • Alex Smola
  • Guy L Steele
  • Namrata Vaswani
چکیده

A (Stochastic) EM in General Expectation-Maximization (EM) is an iterative method for finding the maximum likelihood or maximum a posteriori (MAP) estimates of the parameters in statistical models when data is only partially, or when model depends on unobserved latent variables. This section is inspired from lecture of Dr Namrata Vaswani available at http://www.ece.iastate.edu/∼namrata/EE527 Spring08/emlecture.pdf. We derive EM algorithm for a very general class of model. Let us define all the quantities of interest.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Stochastic bounds for a single server queue with general retrial times

We propose to use a mathematical method based on stochastic comparisons of Markov chains in order to derive performance indice bounds‎. ‎The main goal of this paper is to investigate various monotonicity properties of a single server retrial queue with first-come-first-served (FCFS) orbit and general retrial times using the stochastic ordering techniques‎.

متن کامل

A wavelet method for stochastic Volterra integral equations and its application to general stock model

In this article,we present a wavelet method for solving stochastic Volterra integral equations based on Haar wavelets. First, we approximate all functions involved in the problem by Haar Wavelets then, by substituting the obtained approximations in the problem, using the It^{o} integral formula and collocation points then, the main problem changes into a system of linear or nonlinear equation w...

متن کامل

Fast Learning of On-line EM Algorithm

In this article, an on-line EM algorithm is derived for general Exponential Family models with Hidden variables (EFH models). It is proven that the on-line EM algorithm is equivalent to a stochastic gradient method with the inverse of the Fisher information matrix as a coeecient matrix. As a result, the stochastic approximation theory guarantees the convergence to a local maximum of the likelih...

متن کامل

The Effects of Monetary Policy on Output and Inflation in Afghanistan: A Dynamic Stochastic General Equilibrium Approach

T he consequences of the decisions and policies taken for the reconstruction of Afghanistan during the past decade, showed that there should be paid more attention to the monetary policies. There is also a question that whether monetary policies have the potential to affect the production and inflation in Afghanistan or not. The aim of this paper is to explore this effect by designin...

متن کامل

Optimal Stochastic Control in Continuous Time with Wiener Processes: General Results and Applications to Optimal Wildlife Management

We present a stochastic optimal control approach to wildlife management. The objective value is the present value of hunting and meat, reduced by the present value of the costs of plant damages and traffic accidents caused by the wildlife population. First, general optimal control functions and value functions are derived. Then, numerically specified optimal control functions and value func...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2016