Estimating Weibull Parameters Using Least Squares and Multilayer Perceptron vs. Bayes Estimation

The Weibull distribution is regarded as among the finest in family of failure distributions. One most commonly used parameters (WD) ordinary least squares (OLS) technique, which useful reliability and lifetime modeling. In this study, we propose an approach based on multilayer perceptron (MLP) neural network called OLSMLP that resilience OLS method. MLP solves problem heteroscedasticity distorts estimation WD due to presence outliers, eases difficulty determining weights case weighted square (WLS). Another method proposed by incorporating a weight into general entropy (GE) loss function estimate obtain modified (WGE). Furthermore, Monte Carlo simulation performed examine performance comparison with approximate Bayesian (BLWGE) using GE function. results showed two methods produced good estimates even for small sample sizes. addition, techniques here are typically preferred options when estimating compared other available methods, terms mean squared error requirements related time.