Abstract: A method for advanced condition monitoring of an asset system includes sensing actual values of an operating condition for an operating regime of the asset system using at least one sensor; estimating sensed values of the operating condition by using an auto-associative neural network; determining a residual vector between the estimated sensed values and the actual values; and performing a fault diagnostic on the residual vector. In another method, an operating space of the asset system is segmented into operating regimes; the auto-associative neural network determines estimates of actual measured values; a residual vector is determined from the auto-associative neural network; a fault diagnostic is performed on the residual vector; and a change of the operation of the asset system is determined by analysis of the residual vector. An alert is provided if necessary. A smart sensor system includes an on-board processing unit for performing the method of the invention.

Abstract: A method of generating an indication of a probability of a hypothesis being correct based on a set of observations includes obtaining data representing first and second sets of observations. Data representing a set of hypotheses at least partially derivable from the first and the second set of observations can also be obtained. Plural data associations can be generated between at least some data in the first and second sets to indicate a probability of at least some of the generated data associations being correct. The data representing the set of hypotheses and the indication of the probability of at least some of the generated data associations being correct can be used to generate an indication of a probability of at least one of the hypotheses represented by the data being correct.

Abstract: A method and apparatus for estimating velocity in a subsurface region. Seismic data for a subsurface region may be received. One or more attributes for the seismic data may be calculated. A posterior distribution may be generated. The posterior distribution may represent one or more probabilities of one or more velocities for the attributes. A velocity with uncertainty may be determined for the subsurface region based on the posterior distribution. A pore pressure with uncertainty may be determined based on the velocity with uncertainty.

Abstract: Metrology data from a semiconductor treatment system is transformed using multivariate analysis. In particular, a set of metrology data measured or simulated for one or more substrates treated using the treatment system is obtained. One or more essential variables for the obtained set of metrology data is determined using multivariate analysis. A first metrology data measured or simulated for one or more substrates treated using the treatment system is obtained. The first obtained metrology data is not one of the metrology data in the set of metrology data earlier obtained. The first metrology data is transformed into a second metrology data using the one or more of the determined essential variables.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining a presumed activity associated with a mobile device. A plurality of sensor values detected by one or more sensors onboard the mobile device is received over a period of time. A plurality of derived values is calculated from the plurality of sensor values. The derived values are selectively combined to generate one or more abstract values. A presumed activity is identified from a plurality of possible activities based on a level of similarity between the one or more abstract values and expected values of each of the plurality of possible activities that correspond to the one or more abstract values.

Abstract: Methods and systems for use in predicting wind turbine failures are provided. One example method includes determining a parametric profile for a component of a wind turbine from operating data of a plurality of wind turbines, determining an anomaly profile for the component of the wind turbine from anomaly alerts from a plurality of wind turbines, and determining a probability of failure for the component of the wind turbine based on the parametric profile and the anomaly profile. The parametric profile defines at least one parametric event associated with the component prior to failure of the component, and the anomaly profile defines at least one anomaly associated with the component prior to failure of the component.

Abstract: Power quality (PQ) events in a waveform in a power distribution network are detected by sampling the waveform. For each sample, a normalized sample is determined by subtracting an undistorted waveform from the sample. An accumulated weighted log-likelihood ratio (LLR) is determined for all normalized samples, as well as a minimum of the accumulated weighted LLR. If the difference between the accumulated weighted LLR an the minimum LLR is greater than a predetermined threshold, then the PQ event is signaled. The network can be a smart grid.

Abstract: A method of absolute nuclear material assay of an unknown source comprising counting neutrons from the unknown source and providing an absolute nuclear material assay utilizing a model to optimally compare to the measured count distributions. In one embodiment, the step of providing an absolute nuclear material assay comprises utilizing a random sampling of analytically computed fission chain distributions to generate a continuous time-evolving sequence of event-counts by spreading the fission chain distribution in time.

Abstract: An operations support system for an engine includes an engine prediction unit configured to receive an engine rating condition and scalars associated with the engine, and to generate condition indicators based on the engine rating condition and the scalars; and a power assurance algorithm unit coupled to the engine prediction unit and configured to generate health indicators based on the condition indicators. The health indicators include a power assurance number.

Abstract: A method and apparatus are described for determining a threshold, including selecting a desired false result probability, and selecting a threshold value from a threshold table corresponding to the desired false result probability. Also described are a method and apparatus for determining a threshold corresponding to a desired false result probability, including calculating a false result probability, calculating a difference value between the calculated false result probability and the desired false result probability, comparing the difference value to a first predetermined value and comparing the difference value to a second predetermined value if a result of the first comparing act is positive.

Abstract: A method of detecting an unhealthy/potentially/failing instrument using apparatus is carried out by measuring a characteristic of the output of an instrument. The measurement of the characteristic is compared to an expected distribution of the instrument when healthy. The probability of the instrument producing such a characteristic measurement, or a value further from the mean of the expected distribution, if it was healthy is calculated. The measured characteristic to an expected distribution of the instrument when unhealthy is compared, and the probability of the instrument producing such a characteristic measurement, or a value further from the mean of the expected distribution, if it was unhealthy is calculated. The probability of the measured characteristic being produced by the instrument when healthy and when unhealthy is compared. A confidence value indicative of the likelihood of the instrument being unhealthy is then produced.

Abstract: Robust methods are developed to provide bounds and probability distributions for the locations of objects as well as for associated variables that affect the accuracy of the location such as the positions of stations, the measurements, and errors in the speed of signal propagation. Realistic prior probability distributions of pertinent variables are permitted for the locations of stations, the speed of signal propagation, and errors in measurements. Bounds and probability distributions can be obtained without making any assumption of linearity. The sequential methods used for location are applicable in other applications in which a function of the probability distribution is desired for variables that are related to measurements.

Abstract: Predicting the probability of detection of major and minor defects in a structure includes simulating a plurality of N defects at random locations in a region specified by an array of transducers. Defect size is incremented until it intersects one path between two transducers. The defect size is again incremented until it intersects two or more adjacent paths between pairs of transducers. The number of major defects up to a selected size is determined by the total number of single path intersections by defects up to the selected size. The number of minor defects up to a selected size is determined on the basis of the total number of defects intersecting two or more paths up to the selected size. The probability of detection up to a selected size is the cumulative number of major or minor defects up to the selected size normalizing by N.

Abstract: Method of evaluating precision of output data using error propagation includes performing numerical differentiation using input data in data processing, and thereby obtaining a Jacobian matrix J of the data processing; estimating variance-covariance of errors of the input data, and thereby obtaining an error matrix D of the input data; and calculating an error matrix from the Jacobian matrix J and the error matrix D of the input data, the error matrix R representing variance-covariance of errors of output data.

Abstract: Methods, program products, computer program products and systems for uniquely identifying an individual within a population to the exclusion of all others within the population by comparing data from unique data sources based on the name of the individual for providing a collection of preliminary suspects. An individual search service provider is then searched for additional data relating to the collection of preliminary suspects to locate any potential matches, which are reconciled with data from the unique data sources to locate at least a portion of a unique identifier that may be associated with the individual. The potential matches and portion of the unique identifier are compared with other data records within the individual search data source for locating a complete unique identifier that may be associated with the individual, followed by determining whether or not this complete unique identifier uniquely identifies the individual.

Abstract: In one embodiment, we describe a method that generates seasonality rules for anomaly detection for a hierarchical/tree based data structure. A new algorithm for processing nodes in hierarchy, as well as business rules for nodes, is described. Variations and examples are given to describe different scopes and embodiments of the invention. Exclusion criteria and children nodes are used as some examples for the implementations, with flow charts to describe the methods of application, as examples.

Abstract: Apparatus and method for improving the performance of a threshold-based detector or classifier, or a generic detector or classifier and increasing the probability of detecting at least one object in an image using novel algorithms and stochastic resonance noise is provided, where a suitable dose of noise is introduced to the image data such that the performance of the above-referenced detectors or classifiers is improved without altering the detector's or classifier's parameters. Several stochastic resonance (SR) noise-based detection and classification enhancement schemes are presented. The SR noise-enhanced detection and classification schemes can improve any algorithms and systems. To implement these schemes, the only knowledge that is needed is the original input data (no matter 1D, 2D, 3D or others) and the output (detection results) of the existing algorithms and systems.

Abstract: Estimating monthly heating oil consumption of a building that uses heating oil and non-oil source of energy, may include separating by applying statistical models, yearly consumption of oil data associated with the building into base load oil consumption and space heating oil consumption. The separating may also include determining monthly base load oil consumption associated with the building. Monthly space heating consumption of oil may be estimated by applying a heating degree day density function to the space heating oil consumption. The monthly space heating consumption may be aggregated with the monthly base load oil consumption to estimate the monthly heating oil consumption.

Abstract: Beamed data can be obtained from shot gather data, and debeamed data can be obtained from beamed data. Shot gather data for a geophysical volume of interest in a first domain and/or beamed data in a third domain may be received. The first domain has a component relating to seismic wave propagation time and a spatial component relating to lateral spacing. The third domain has a component relating to local plane wave arrival time at a beam center surface location and a component relating to plane wave arrival directions. Data may be transformed between the first domain and a second domain. The second domain has a frequency component and a spatial component corresponding to the spatial component of the first domain. A forward or reverse transform may be applied to transform shot gather data in the first domain to/from beamed data in the third domain.

Abstract: Methods and apparatus for enhancing performance curve generation, damage monitoring, and improving non-destructive testing performance. Damage standards used for performance curve generation are monitored using a non-destructive testing (NDT) sensor during a damage evolution test performed with the standard. The evolution test may be intermittently paused to permit ground truth data to be collected in addition to the NDT sensor data. A damage evolution model may be used to estimate ground truth data during the intervening periods of the damage evolution test. The NDT sensor data and ground truth data are used to generate performance curves for the NDT system. Multiple sensors may be monitored at multiple locations on the damage standard and multiple damage evolution tests may be performed with multiple damage standards.

Abstract: The adaptive contrast enhancer uses an adaptive histogram equalization-based approach to improve contrast in a video signal. For each video frame, the histogram of the pixel luminance values is calculated. The calculated histogram is divided into three regions that are equalized independently of the other. The equalized values are averaged with the original pixel values with a weighting factor that depends on the shape of the histogram. The weighting factors can be also chosen differently for the three regions to enhance the darker regions more than the brighter ones. The statistics calculated from one frame are used to enhance the next frame such that frame buffers are not required. Many of the calculations are done in the inactive time between two frames.

Abstract: Various embodiments of the present invention provide systems and methods for data processing. As an example, a data processing circuit is disclosed that includes: a data detector circuit, a calibration circuit, and an enable circuit. The data detector circuit is operable to apply a data detection algorithm to a data input to yield a detected output based at least in part on control values. The calibration circuit operable to update the control values based at least in part on the data input, the detected output, and a calibration circuit enable. The calibration circuit enable is generated by the enable circuit based at least in part on the detected output.

Abstract: In particular embodiments, fusing multi-sensor data sets includes receiving a first sensor data set and a second sensor data set generated in response to sensing a structure. The sensor data sets describe structural features of the structure. First delta vector sets are generated for the first sensor data set, and second delta vector sets are generated for the second sensor data set. Each delta vector set comprises delta vectors indicating relative geometrical relationships between a structural feature and other structural features. Association scores are determined for delta pairs comprising a first delta vector set and a second delta vector set. Same feature delta pairs are identified according to the association scores. A same feature delta vector set comprises a delta pair corresponding to the same structural feature.

Abstract: A method and system for identifying unknown illicit networks uses node data, link data, and network data in a recursive analysis that computes node probabilities by combining patterns based on nodes, links and the topology of the network structure present simultaneously in the data. An iterative balancing algorithm is used to make the probability values self-consistent.

Abstract: A method for generating scatter-plots similar to classic ROC curves to gain an understanding of the tradeoffs between probability of detection and probability of false alarm for detections systems. The method uses a stochastic optimization and an objective function to maximize probabilities of detection for probabilities of false alarm that are arbitrarily close to some target probability of false alarm.

Abstract: Operational parameters of a single-photon detector are determined with a source of photon pairs. At a fixed trigger, a dark count probability and a series of count probabilities at different optical powers are determined. A particular optical power is selected by adjusting the optical power of the source of photon pairs. The dark count probability and the count probabilities are determined for different trigger rates. The operational parameters include efficiency, afterpulsing constant, and detrap time. The operational parameters are computed by fitting the computed dark count probabilities and count probabilities to a user-defined relationship.

Abstract: A method of establishing statistically valid assay means and ranges for quality control materials, used to qualify medical testing machines, utilizes tests on a new lot of quality control material to establish an assay mean, and uses data from a database of historical test results to establish an assay range. The system may estimate the variability of test results from prior lot data, and then compute the limits of the assay range such that a new test on a new lot of the quality control material will be expected to fall within the range with a specified probability. Because historical data is used to estimate the test variability, the number of new tests required to specify a statistically valid mean and range may be dramatically reduced, as compared with establishing the mean and range based only on tests of the new lot of material.

Abstract: The present invention provides automatic gating methods that are useful to gate populations of interest in multidimensional data, wherein the populations of interest are only a subset of the populations identifiable in the data. The populations are modeled as a finite mixture of multivariate probability distributions, preferably normal or t distributions. The distribution parameters that provide a best fit of the model distribution to the data are estimated using an Expectation Maximization (EM) algorithm that further includes a dynamic neighborhood thresholding that enables gating of a subset of the clusters present in the data.

Abstract: This invention concerns remote sensor networks, and particularly energy management for wireless sensor networks. In a first aspect the invention is a wireless sensor node specified to operate for a given lifetime, comprising: an onboard computer system and a set of one or more associated sensors. The computer system operates to periodically sample data from each sensor of the set of associated sensors, and to store a multi-state model representing one or more phenomena described by the collected data. And, the computer system operates to calculate a value associated with movement of the phenomena between the states of the multi-state model, and to adjust the rate of sampling of one or more of the set of associated sensors depending on the calculated value. In other aspects the invention is a network of sensor nodes and a method of operation.

Abstract: There is provided an electronic apparatus system. The system includes: an electronic apparatus; and a server configured to receive usage information about usage status of the electronic apparatus. The server includes: a first database; a second database; a writer configured to store the usage information in the first database and to write a flag value into the second database, wherein the flag value indicates whether or not the usage information is transmitted from the electronic apparatus through a network; a generator configured to generate a first parameter for calculating a failure probability of the electronic apparatus, based on the usage information and repair information of the electronic apparatus, wherein the repair information is stored in the server in advance; and a transmitter configured to transmit the first parameter to the electronic apparatus.

Abstract: A method and system of predicting a maintenance schedule and estimating a cost for warranty service of systems, for example, hardware systems, is provided. The method and system in one aspect may collect component usage data and obtain component reliability data for each of the plurality of components. A component life distribution module (CLDM) may be constructed based on the real time component usage data and the component reliability data, and from CLDM a mean time to fail data may be obtained. A system life distribution model (SLDM) may be selected and the mean time to fail data input into the SLDM to obtain an operating environment index that represents the effect of environment conditions on the lifetime and reliability. A maintenance schedule may be predicted and an adjusted cost of warranty service may be calculated based on the operating environment index for each of the plurality of components.

Abstract: A method of statistical analysis is based on an accelerated degradation model for estimating the failure rate from a set of accelerated life data (e.g. stress, humidity, temperature, voltage, resistance, vibration, etc.). The method re-samples randomly distributed data and organizes it into subsets that may be ordered. The maximum is found for each subset. From the maximums a parametric LEV distribution is determined. Maximum Likelihood Estimation methods may be used to find the parameters employed in the determined LEV distribution. The failure rate is calculated using the determined LEV distribution of the sample maximums.

Abstract: The disclosed method includes: calculating a first expected value of the number of failures for each combination of a feature that is a failure factor and a first group regarding classification elements of first semiconductor devices for which a failure is analyzed and second semiconductors on which a same circuit as the first semiconductors is implemented, from first data for each first group and a predetermined expression, wherein the first data includes the number of actual failures occurred in the first group and first feature values of features; and calculating, for each feature, a first indicator value representing similarity between a distribution of the first expected values over the first groups and a distribution of the numbers of actual failures over the first groups, from the first expected value for each combination of the feature and the first group and the number of actual failures for each first group.

Abstract: Disclosed herein are methods for monitoring a thermoelectric heating and cooling device (103, 203, 303) of a system (100, 200, 300) for cycling liquid reaction mixtures through a series of temperature excursions, and such systems adapted to perform these methods. A first quantity being either an electric current or an electric voltage is applied to the heating and cooling device (103, 203) or a portion (314) thereof (303), and a second quantity being either the non-selected first quantity or temperature is measured to obtain a first test value. The selected first quantity is applied to another thermoelectric heating and cooling device (104, 204) or another portion (315) of the first device (303), and the second quantity is measured to obtain a second test value. Comparison of the first and second test values yields a monitoring value which is compared with a pre-defined threshold value to obtain a monitoring result.

Abstract: There is provided a deterministic component model identifying apparatus for determining a type of a deterministic component contained in a probability density function supplied thereto.

Abstract: According to one embodiment, a first reading unit reads, from a defect database, first identification information corresponding to a first product type. A second reading unit reads, from a monitoring database, defect monitoring information corresponding to the first identification information and non-defect monitoring information that corresponds to the first product'type and that is other than the defect monitoring information. A generating unit generates a defect, model that models a probability of products becoming defective within a predetermined time period with respect to the monitoring information, based on the defect monitoring information and the non-defect monitoring information. A first calculating unit calculates a defect probability of products of a second product type by inputting the monitoring information corresponding to the second product type into the defect model. A second calculating unit calculates a defect rate of the products of the second product type based on the defect probability.

Abstract: A system for determining occurrence factors of particles includes a user interface device, and an apparatus for detecting the occurrence factors of particles. The apparatus for detecting the occurrence factors of particles includes a storage unit that stores a program for executing a calculation method for calculating a likelihood of each of the occurrence factors of particles in the form of a score; and a calculation unit for calculating the score for each of the occurrence factors of particles based on particle distributions at least on a surface of a substrate using the stored program. The user interface device displays the calculated score for each of the occurrence factors of particles.

Abstract: A signal transmission system evaluation apparatus acquires statistics about a variation in a characteristic value and a limit value of the characteristic value corresponding to a given range of variation, with respect to each of the characteristic values which represent characteristics of the components. The apparatus calculates a probability distribution with respect to each of the characteristic values, based on the statistic acquired, calculates an eye-opening of the signal transmission system in case that the characteristic value is the limit value, makes an adjustment of the limit value. The apparatus calculates a yield rate of the signal transmission system based on the probability distribution and the limit value.

Abstract: A method, system and computer program product for distinguishing between a sensor fault and a process fault in a physical system and use the results obtained to update the model. A Bayesian network is designed to probabilistically relate sensor data in the physical system which includes multiple sensors. The sensor data from the sensors in the physical system is collected. A conditional probability table is derived based on the collected sensor data and the design of the Bayesian network. Upon identifying anomalous behavior in the physical system, it is determined whether a sensor fault or a process fault caused the anomalous behavior using belief values for the sensors and processes in the physical system, where the belief values indicate a level of trust regarding the status of its associated sensors and processes not being faulty.

Abstract: A method and system for automatically determining an optimal re-training interval for a fault diagnoser based on online monitoring of the performance of a classifier are disclosed. The classifier generates a soft measure of membership in association with a class based on a training data. The output of the classifier can be utilized to assign a label to new data and then the members associated with each class can be clustered into one or more core members and potential outliers. A statistical measure can be utilized to determine if the distribution of the outliers is sufficiently different than the core members after enough outliers have been accumulated. If the outliers are different with respect to the core members, then the diagnoser can be re-trained; otherwise, the output of the classifier can be fed to the fault diagnoser.

Abstract: A prediction device for predicting a prediction interval being a time interval predicted with a high possibility of appliance utilization starts includes: an evaluation unit calculating, using utilization interval history data, an evaluation value for determining a prediction scheme for each appliance; a determination unit, based on the evaluation value at least for each appliance, determining at least one of a first and a second prediction scheme as the scheme of the prediction interval for the appliance, the first prediction scheme using a property that the appliance is utilized in a predetermined cycle, and the second prediction scheme using a property that the utilization of a second appliance being the appliance is started in a period from a start or an end of utilization of a first appliance being another appliance to a passage of a predetermined time; a prediction unit predicting a prediction interval according to the determined scheme.

Abstract: According to one embodiment, an existent person count estimation apparatus includes motion sensors and following units. The collection unit generates human sensing information. The instance prediction unit predicts second instances from the first instances by using the transition matrix. The likelihood calculation unit calculates likelihoods of the second instances using the time information items. The instance selection unit selects one or more third instances having likelihoods higher than a threshold. The output unit generates output information including estimate values of existent person counts for the first areas included in the third instances.

Abstract: A method and system for partitioning (clustering) large amounts of data in a relatively short processing time. The method involves providing a first data matrix and a second data matrix where each of the first and second data matrices includes one or more variables, and a plurality of data points. The method also involves determining a first score from the first data matrix using a partial least squares (PLS) analysis or orthogonal PLS (OPLS) analysis and partitioning the first and second data matrices (e.g., row-wise) into a first group and a second group based on the sorted first score, the variance of the first data matrix, and a variance of the first and second groups relative to the variances of the first and second data matrices.

Abstract: The invention relates to a method and apparatus for characterizing the imperfections of the skin. The apparatus comprises: a) a digital camera allowing the taking of at least one digital image of at least one determined skin zone, the said image being defined by a multiplicity of pixels, that is transmitted to a digital image processing device; b) means of splitting the digital image into three color planes: red, green, blue, termed R, G, B, with the aid of the said image processing device; c) means of extracting a single of these planes; d) means of calculating at least one of the following parameters on the basis of this plane: the mean grey level of the image; the surface area of the imperfections of the skin of the said image, some of which are related to aging, such as wrinkles, lines, spots, dark circles, pigmentary imperfections, slackening or loosening zones; the variance of the grey levels over the set of pixels of the image as parameter representative of the homogeneity of the skin.

Abstract: A method includes passing a lot through a production process and evaluating a statistical quality of the production process. Additionally, the method includes calculating an advanced process control (APC) recipe parameter adjustment (RPA) distribution value and determining if sampling is indicated. Furthermore, the method includes, if sampling is indicated, performing a measurement process of the lot.

Abstract: Embodiments include systems and methods of non-coherent signal processing. For example, one embodiment includes a method of a non-coherent signal detection. The method includes receiving data indicative of a received signal, and evaluating a probability density function indicative of a signal parameter based on the received data. Evaluating the probability density function comprises identifying at least one of a plurality of terms of summation having at least one peak of the probability density function and evaluating the plurality of terms of a non-central chi-square distribution determining the signal parameter based at least in part on the received data and the calculated probability density.

Abstract: The EMERGENCY RESPONSE MANAGEMENT APPARATUSES, METHODS AND SYSTEMS (“ERMS”) transform emergency related inputs and sensor information into a threat indication category, which is distributed to individuals and/or first responders for managing the threat. In one implementation, the method includes an emergency management processor-implemented method that receives sensor readings from one or more sensor devices and generates risk factors for the at least one sensor device. The generated risk factors are then curve fitted to a plurality of statistical distribution curves including both non-extreme and extreme statistical distributions, wherein each of the statistical distribution curves is indicative of a threat category. The threat category is then determined based on the generated risk factors that provide a best fit with one of the plurality of statistical distribution curves.

Abstract: Embodiments of the invention disclose a system and a method for determining a pose of a probe relative to an object by probing the object with the probe, comprising steps of: determining a probability of the pose using Rao-Blackwellized particle filtering, wherein a probability of a location of the pose is represented by a location of each particle, and a probability of an orientation of the pose is represented by a Gaussian distribution over orientation of each particle conditioned on the location of the particle, wherein the determining is performed for each subsequent probing until the probability of the pose concentrates around a particular pose; and estimating the pose of the probe relative to the object based on the particular pose.

Abstract: System and method for implementing wafer acceptance test (“WAT”) advanced process control (“APC”) are described. In one embodiment, the method comprises performing an inter-metal (“IM”) WAT on a plurality of processed wafer lots; selecting a subset of the plurality of wafer lots using a lot sampling process; and selecting a sample wafer group using the wafer lot subset, wherein IM WAT is performed on wafers of the sample wafer group to obtain IM WAT data therefore. The method further comprises estimating final WAT data for all wafers in the processed wafer lots from IM WAT data obtained for the sample wafer group and providing the estimated final WAT data to a WAT APC process for controlling processes.

Abstract: For an integrated circuit associated with a plurality of parameters whose values are described by a first probability distribution function, a method for estimating a failure probability includes selecting a first plurality of samples, performing a first test to determine an outcome for each of the first plurality of samples and identifying failed samples, and clustering the failed samples using a computer-implemented cluster forming method that, in some cases, returns multiple clusters. The method also includes forming a probability distribution function for each of the clusters, forming a composite probability distribution function that includes a weighted combination of the first probability distribution function and the probability distribution function for each of the clusters. The method further includes selecting a second plurality of samples using the composite probability distribution function and performing a second test to determine an outcome for each of the second plurality of samples.