Case Studies for Model Efficiency: Special Sampling and MLE Bias Correction
نویسندگان
چکیده
Driven by the rising need for model efficiency to enable timely and accurate data analytics, stochastic modelers in the actuarial community have been seeking innovative alternatives to increase model reality, dynamics, capacity, speed, and precision. These challenges have motivated this paper to demonstrate free research software tools for case studies in financial stochastic modeling. Case studies considered in this paper use two free software applications: (1) CSTEP (Cluster Sampling for Tail Estimation of Probability) for reducing sampling bias error with specialized user-defined distance sampling, and (2) AMOOF2 (Actuarial Model Outcome Optimal Fit Version 2.0) for correcting small-sampling bias from fitting parametric probability density functions via maximum likelihood estimation to a small sample. These two software applications were used to reveal insights regarding the probability distribution of statutory ending surplus for a large business block of life annuity contracts, while comparing special sampling techniques that affect outcomes and model efficiency, assessed by the bias of fitting error.
منابع مشابه
A Numerical Comparison of the Modified Method of Moments and Maximum Likelihood for the Estimation of Short-Range Correlations
This paper focuses on the inference of the correlation parameters from data that have either spatial or temporal short-range dependence. A comparative study of two inference methods is presented: the established method of maximum likelihood estimation (MLE) and the recently proposed modified method of moments (MMoM). For reasons of computational efficiency the comparisons are based on simulatio...
متن کاملGrammatical Error Correction with Neural Reinforcement Learning
We propose a neural encoder-decoder model with reinforcement learning (NRL) for grammatical error correction (GEC). Unlike conventional maximum likelihood estimation (MLE), the model directly optimizes towards an objective that considers a sentence-level, task-specific evaluation metric, avoiding the exposure bias issue in MLE. We demonstrate that NRL outperforms MLE both in human and automated...
متن کامل1 Fourth and Fifth Order Efficiency : Fisher Information
Loss of (Fisher) information, whose concept was de ned by Fisher, is one of important measures of asymptotic e ciency. Based on Fisher information, this paper studies fourth and fth order asymptotic e ciency of estimators in a curved exponential family of distributions with a structural parameter vector. In particular, we show that the bias-corrected MLE with a certain bias correction factor is...
متن کاملFinite-sample Moments of the Mle for the Binary Logit Model
Author Contact: David E., Dept. of Economics, University of Victoria, P.O. Box 1700, STN CSC, Victoria, B.C., Canada V8W 2Y2; e-mail: [email protected]; Voice: (250) 721-8540; FAX: (250) 721-6214 Abstract We examine the finite sample properties of the MLE for the Logit model with random covariates. We derive the second order bias and MSE function for the MLE in this model, and undertake some numer...
متن کاملStatistical Estimation under Length Biased Distributions
In simple random sampling, there is an equal chance of selection for each unit in the population. However, there are some practical situations where this might not be ideal. We consider the case of length-biased sampling, whereby the chance of selection of a unit is proportional to the length of the unit, by which the length-biased density is defined. Specifically we consider the case of a cont...
متن کامل