Adaptive Importance Sampling via Stochastic Convex Programming
نویسندگان
چکیده
We show that the variance of the Monte Carlo estimator that is importance sampled from an exponential family is a convex function of the natural parameter of the distribution. With this insight, we propose an adaptive importance sampling algorithm that simultaneously improves the choice of sampling distribution while accumulating a Monte Carlo estimate. Exploiting convexity, we prove that the method’s unbiased estimator has variance that is asymptotically optimal over the exponential family.
منابع مشابه
Faster Optimization through Adaptive Importance Sampling
The current state of the art stochastic optimization algorithms (SGD, SVRG, SCD, SDCA, etc.) are based on sampling one active datapoint uniformly at random in each iteration. Changing these probabilities to better reflect the importance of each datapoint is a natural and powerful idea. In this thesis we analyze Stochastic Coordinate Descent methods with fixed non-uniform and adaptive sampling. ...
متن کاملStochastic Dynamic Programming with Markov Chains for Optimal Sustainable Control of the Forest Sector with Continuous Cover Forestry
We present a stochastic dynamic programming approach with Markov chains for optimal control of the forest sector. The forest is managed via continuous cover forestry and the complete system is sustainable. Forest industry production, logistic solutions and harvest levels are optimized based on the sequentially revealed states of the markets. Adaptive full system optimization is necessary for co...
متن کاملEscaping the Local Minima via Simulated Annealing: Optimization of Approximately Convex Functions
We consider the problem of optimizing an approximately convex function over a bounded convex set in Rn using only function evaluations. The problem is reduced to sampling from an approximately log-concave distribution using the Hit-and-Run method, which is shown to have the same O∗ complexity as sampling from log-concave distributions. In addition to extend the analysis for log-concave distribu...
متن کاملOptimal Importance Sampling in Securities Pricing
To reduce variance in estimating security prices via Monte Carlo simulation, we formulate a parametric minimization problem for the optimal importance sampling measure, which is solved using in nitesimal perturbation analysis (IPA) and stochastic approximation (SA). Compared with existing methods, the IPA estimator we derive is more universally applicable and more computationally e cient. Under...
متن کاملDecision making in forest management with consideration of stochastic prices
The optimal harvesting policy is calculated as a function of the entering stock, the price state, the harvesting cost, and the rate of interest in the capital market. In order to determine the optimal harvest schedule, the growth function and stumpage price process are estimated for the Swedish mixed species forests. The stumpage price is assumed to follow a stochastic Markov process. A stoch...
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 2015