Elmira Popova

Design of experiments on neural network's training for nonlinear time series forecasting

In this study, the statistical methodology of Design of Experiments (DOE) was applied to better determine the parameters of an Artificial Neural Network (ANN) in a problem of nonlinear time series forecasting. Instead of the most common trial and error technique for the ANNs training, DOE was found to be a better methodology. The main motivation for this study was to forecast seasonal nonlinear time series-that is related to many real problems such as short-term electricity loads, daily prices and returns, water consumption, etc. A case study adopting this framework is presented for six time series representing the electricity load for industrial consumers of a production company in Brazil.

Renewal reward processes with fuzzy rewards and their applications to T-age replacement policies

The application of fuzzy set theory to renewal reward processes is proposed in this paper. The reward is modeled as a fuzzy random variable. A theorem which presents the long-run average fuzzy reward per unit time is stated. A procedure to obtain the best T-age replacement policy with fuzzy cost structure is developed. The original problem is transformed into a nonlinear program in order to evaluate the membership of the long-run average fuzzy cost per unit time.

Maintenance policies with two-dimensional warranty

Abstract A new maintenance policy which minimizes the total expected servicing cost for an item with two-dimensional warranty is proposed. An iterative procedure to estimate the items failure rate function from historical observations and an optimization algorithm based on Monte Carlo simulation are applied to obtain the best maintenance policy. Numerical examples and sensitivity analysis are presented.

Algorithmic guided screening of drug combinations of arbitrary size for activity against cancer cells

The standard treatment for most advanced cancers is multidrug therapy. Unfortunately, combinations in the clinic often do not perform as predicted. Therefore, to complement identifying rational drug combinations based on biological assumptions, we hypothesized that a functional screen of drug combinations, without limits on combination sizes, will aid the identification of effective drug cocktails. Given the myriad possible cocktails and inspired by examples of search algorithms in diverse fields outside of medicine, we developed a novel, efficient search strategy called Medicinal Algorithmic Combinatorial Screen (MACS). Such algorithms work by enriching for the fitness of cocktails, as defined by specific attributes through successive generations. Because assessment of synergy was not feasible, we developed a novel alternative fitness function based on the level of inhibition and the number of drugs. Using a WST-1 assay on the A549 cell line, through MACS, we screened 72 combinations of arbitrary size formed from a 19-drug pool across four generations. Fenretinide, suberoylanilide hydroxamic acid, and bortezomib (FSB) was the fittest. FSB performed up to 4.18 SD above the mean of a random set of cocktails or “too well” to have been found by chance, supporting the utility of the MACS strategy. Validation studies showed FSB was inhibitory in all 7 other NSCLC cell lines tested. It was also synergistic in A549, the one cell line in which this was evaluated. These results suggest that when guided by MACS, screening larger drug combinations may be feasible as a first step in combination drug discovery in a relatively small number of experiments. [Mol Cancer Ther 2009;8(3):521–32]

A Bayesian stochastic programming approach to an employee scheduling problem

Bayesian forecasting models provide distributional estimates for random parameters, and relative to classical schemes, have the advantage that they can rapidly capture changes in nonstationary systems using limited historical data. Unlike deterministic optimization, stochastic programs explicitly incorporate distributions for random parameters in the model formulation, and thus have the advantage that the resulting solutions more fully hedge against future contingencies. In this paper, we exploit the strengths of Bayesian prediction and stochastic programming in a rolling-horizon approach that can be applied to solve real-world problems. We illustrate the methodology on an employee production scheduling problem with uncertain up-times of manufacturing equipment and uncertain production rates. Computational results indicate the value of our approach.

Network deployment of radiation detectors with physics-based detection probability calculations

We describe a model for deploying radiation detectors on a transportation network consisting of two adversaries: a nuclear-material smuggler and an interdictor. The interdictor first installs the detectors. These installations are transparent to the smuggler, and are made under an uncertain threat scenario, which specifies the smuggler’s origin and destination, the nature of the material being smuggled, the manner in which it is shielded, and the mechanism by which the smuggler selects a route. The interdictor’s goal is to minimize the probability the smuggler evades detection. The performance of the detection equipment depends on the material being sensed, geometric attenuation, shielding, cargo and container type, background, time allotted for sensing and a number of other factors. Using a stochastic radiation transport code (MCNPX), we estimate detection probabilities for a specific set of such parameters, and inform the interdiction model with these estimates.

Prioritizing Project Selection

We consider capital investments under uncertainty. A typical approach to this problem, when the problem parameters are assumed known, is via a multi-knapsack model. This model takes as input annual budgets as well as the cost streams and profit—i.e., net present value (NPV)—of each project. Its output is a portfolio of projects with the highest total NPV, observing yearly budget constraints. We argue that such a portfolio fails to hedge against uncertainties in the budgets, the cost streams, and the profits. As an alternative, we propose a model that forms an optimal priority list of projects, incorporating multiple scenarios for these input parameters. We apply our approach to two sets of example projects from the South Texas Project Nuclear Operating Company.

BAYESIAN MAINTENANCE POLICIES DURING A WARRANTY PERIOD

Most of the brand new items are released on the market with a certain type of warranty. A fixed length warranty period is assumed in this paper. A set of maintenance policies which consist of minimal repair and preventive maintenance is analyzed for the case of known and unknown failure parameters of the items lifetime distribution. For the second case, two types of Bayesian policies are considered. An extensive simulation study comparing the performance of these maintenance policies is performed

Group replacement policies for parallel systems whosecomponents have phase distributed failure times

Consider a system of components operating in parallel. Downtime costs are incurred when failed components are not repaired or replaced. There are also fixed, unit repair and replacement costs associated with the system. The failure distributions of the components are assumed to be identically distributed random variables. Results on calculating the expected cost and variance per unit time of various group replacement policies will be provided. Consideration of variance is important since, in many cases, practitioners wish not only to achieve small expected cost but also to reduce variability from cycle to cycle. Phase distributions allow for the modeling of a wide range failure time behavior. Closed‐form results are derived for the three major classes of group replacement policy (m‐failure, T‐age, and(m, T)) when the underlying distribution is of phase type.

Adaptive stochastic manpower scheduling

Bayesian forecasting models provide distributional estimates for random parameters, and relative to classical schemes, have the advantage that they can rapidly capture changes in nonstationary systems using limited historical data. Stochastic programs, unlike deterministic optimization models, explicitly incorporate distributions for random parameters in the model formulation, and thus have the advantage that the resulting solutions more fully hedge against future contingencies. We exploit the strengths of Bayesian prediction and stochastic programming in a rolling-horizon approach that can be applied to solve real-world problems. We illustrate the methodology on an employee scheduling problem with uncertain up-times of manufacturing equipment and uncertain production rates.