Abstract: A computing device trains a fair prediction model. A prediction model is trained and executed with observation vectors. A weight value is computed for each observation vector based on whether the predicted target variable value of a respective observation vector of the plurality of observation vectors has a predefined target event value. An observation vector is relabeled based on the computed weight value. The prediction model is retrained with each observation vector weighted by a respective computed weight value and with the target variable value of any observation vector that was relabeled. The retrained prediction model is executed. A conditional moments matrix is computed. A constraint violation matrix is computed. Computing the weight value through computing the constraint violation matrix is repeated until a stop criterion indicates retraining of the prediction model is complete. The retrained prediction model is output.

Abstract: A bladder fullness monitoring system includes a controller and an active optical sensor that is affixed to a patient's bladder. The sensor emits light onto the bladder and further detects light reflected from the bladder, in order to generate an output signal that indicates an amount of emitted light was reflected back to the detector. The controller is coupled to the optical sensor to receive and interpret the output signals, e.g., to determine when the bladder is full. The controller may be operatively coupled to a urinary control apparatus which uses the output signals to trigger urination in patients who have lost the ability to voluntarily urinate. Embodiments are particularly useful for monitoring bladder fullness in patients who have lost bladder sensation and/or the ability to voluntary urinate and rely on a urinary control apparatus in order to urinate.

Abstract: A computing device trains a fair prediction model. A prediction model is trained and executed with observation vectors. A weight value is computed for each observation vector based on whether the predicted target variable value of a respective observation vector of the plurality of observation vectors has a predefined target event value. An observation vector is relabeled based on the computed weight value. The prediction model is retrained with each observation vector weighted by a respective computed weight value and with the target variable value of any observation vector that was relabeled. The retrained prediction model is executed. A conditional moments matrix is computed. A constraint violation matrix is computed. Computing the weight value through computing the constraint violation matrix is repeated until a stop criterion indicates retraining of the prediction model is complete. The retrained prediction model is output.

Abstract: A computing device trains a fair machine learning model. A predicted target variable is defined using a trained prediction model. The prediction model is trained with weighted observation vectors. The predicted target variable is updated using the prediction model trained with weighted observation vectors. A true conditional moments matrix and a false conditional moments matrix are computed. The training and updating with weighted observation vectors are repeated until a number of iterations is performed. When a computed conditional moments matrix indicates to adjust a bound value, the bound value is updated based on an upper bound value or a lower bound value, and the repeated training and updating with weighted observation vectors is repeated with the bound value replaced with the updated bound value until the conditional moments matrix indicates no further adjustment of the bound value is needed. A fair prediction model is trained with the updated bound value.

Abstract: A bladder fullness monitoring systems includes a controller and an active optical sensor that is affixed to a patient's bladder. The sensor emits light onto the bladder and further detects light reflected from the bladder, in order to generate an output signal that indicates an amount of emitted light was reflected back to the detector. The controller is coupled to the optical sensor to receive and interpret the output signals, e.g., to determine when the bladder is full. The controller may be operatively coupled to a urinary control apparatus which uses the output signals to trigger urination in patients who have lost the ability to voluntarily urinate. Embodiments are particularly useful for monitoring bladder fullness in patients who have lost bladder sensation and/or the ability to voluntary urinate and rely on a urinary control apparatus in order to urinate.

Abstract: An apparatus for locating a source of vibrational cardiac energy within a human, includes a data processing system configured to accept vibrational cardiac data collected by at least three transducers during a plurality of heart cycles. A computer readable medium containing executable computer program instructions which, when executed by the data processing system, cause the data processing system to perform steps that include, performing a time-to-frequency transformation on the vibrational cardiac data acquired within at least a portion of the heart cycles. The time-to-frequency transformation is applied to the vibrational cardiac data acquired synchronously from each transducer to obtain a plurality of complex frequency spectra. The steps include identifying a vernier band (VB) of frequency associated with a feature derived from the plurality of complex frequency spectra.

Abstract: A computing device trains a fair machine learning model. A predicted target variable is defined using a trained prediction model. The prediction model is trained with weighted observation vectors. The predicted target variable is updated using the prediction model trained with weighted observation vectors. A true conditional moments matrix and a false conditional moments matrix are computed. The training and updating with weighted observation vectors are repeated until a number of iterations is performed. When a computed conditional moments matrix indicates to adjust a bound value, the bound value is updated based on an upper bound value or a lower bound value, and the repeated training and updating with weighted observation vectors is repeated with the bound value replaced with the updated bound value until the conditional moments matrix indicates no further adjustment of the bound value is needed. A fair prediction model is trained with the updated bound value.

Abstract: A computing device trains a fair machine learning model. A prediction model is trained to predict a target value. For a number of iterations, a weight vector is computed using the bound value based on fairness constraints defined for a fairness measure type; a weight value is assigned to each observation vector based on the target value and a sensitive attribute value; the prediction model is trained with each weighted observation vector to predict the target value; and a conditional moments vector is computed based on the fairness constraints and the target and sensitive attribute values. Conditional moments difference values are computed. When the conditional moments difference values indicate to adjust the bound value, the bound value is updated and the process is repeated with the bound value replaced with the updated bound value until the conditional moments difference values indicate no further adjustment of the bound value is needed.

Abstract: A method of locating a source of vibrational cardiac energy within a human, includes acquiring time synchronized vibrational cardiac data noninvasively from a surface of the human's body with at least three transducer channels. A time-to-frequency transformation is performed on the vibrational cardiac data to obtain a sequence of heart cycle ensemble averaged power spectral density (PSD) estimates. A feature is identified in the PSD estimates. A vernier band (VB) in frequency is established for the feature. The VB portion of the PSI) estimates that correspond to the feature are processed to obtain a vibrational energy intensity level estimate at a spatial location within the human's body. A metric of clinical significance is obtained from the strength of the estimated vibration intensity within a VB and at a location relative to a background noise level.

Abstract: Systems and methods for machine learning, models, and related explainability and interpretability are provided. A computing device determines a contribution of a feature to a predicted value. A feature computation dataset is defined based on a selected next selection vector. A prediction value is computed for each observation vector included in the feature computation dataset using a trained predictive model. An expected value is computed for the selected next selection vector based on the prediction values. The feature computation dataset is at least a partial copy of a training dataset with each variable value replaced in each observation vector included in the feature computation dataset based on the selected next selection vector. Each replaced variable value is replaced with a value included in a predefined query for a respective variable. A Shapley estimate value is computed for each variable.

Abstract: Embodiments of the invention provide bladder fullness monitoring systems and associated methods for such monitoring. One includes a controller and an active optical sensor that is affixed to a patient's bladder. The sensor emits light onto the bladder and further detects light reflected from the bladder, in order to generate an output signal that indicates an amount of emitted light was reflected back to the detector. The controller is coupled to the optical sensor to receive and interpret the output signals, e.g., to determine when the bladder is full. The controller may be operatively coupled to a urinary control apparatus which uses the output signals to trigger urination in patients who have lost the ability to voluntarily urinate. Embodiments are particularly useful for monitoring bladder fullness in patients who have lost bladder sensation and/or the ability to voluntary urinate and rely on a urinary control apparatus in order to urinate.

Abstract: Method and apparatuses are described to evaluate a state of health of a coronary artery. Vibrational cardiac data from a transducer is received from a first location. The transducer measures vibration of a surface of a human's body non-invasively. The vibrational cardiac data is processed in a second location to create processed vibrational cardiac data. The processing includes separating an unwanted coronary event from the vibrational cardiac data. The processing further includes performing a time-to-frequency transformation on the vibrational cardiac data from a portion of a diastolic interval within a heart cycle to obtain a vibrational frequency power spectrum estimate. The processing includes evaluating a state of health of the coronary artery using the processed vibrational cardiac data.

Abstract: Methods and apparatuses are described to obtain cardiac data, which includes acquiring vibrational field cardiac data from a transducer wherein the transducer measures vibration over a surface of a human's body. An unwanted coronary event is separated from vibrational cardiac data. A transient event is extracted from the vibrational cardiac heart cycle data. The transient event occurs during a diastolic interval within a heart cycle. The transient event is evaluated for a condition of coronary artery blood flow turbulence and a condition of health of a coronary artery is assessed from a feature in the vibrational frequency power spectrum estimate.

Abstract: Methods and apparatuses are described to obtain cardiac data, which includes acquiring vibrational field cardiac data from a transducer wherein the transducer measures vibration over a surface of a human's body. An unwanted coronary event is separated from vibrational cardiac data. A transient event is extracted from the vibrational cardiac heart cycle data. The transient event occurs during a diastolic interval within a heart cycle. The transient event is evaluated for a condition of coronary artery blood flow turbulence and a condition of health of a coronary artery is assessed from a feature in the vibrational frequency power spectrum estimate.

Abstract: A cross arm is provided, for use in a support structure for conductors within an electrical grid. The cross arm is made of weather resistant glass fibre reinforced two component polyurethane based polymers and has a “C” cross sectional shape typically arranged in a double configuration. Pairs of the cross arms are attached to utility poles in a parallel position on opposite sides of the utility pole to meet predefined design conditions. Alternatively, a single cross arm may be attached to a utility pole.

Abstract: A cross arm is provided, for use in a support structure for conductors within an electrical grid. The cross arm is made of weather resistant glass fiber reinforced two component polyurethane based polymers and has a “C” cross sectional shape typically arranged in a double configuration. Pairs of the cross arms are attached to utility poles in a parallel position on opposite sides of the utility pole to meet predefined design conditions. Alternatively, a single cross arm may be attached to a utility pole.

Abstract: A method includes acquiring vibrational cardiac data from an array of N transducers wherein the transducers are coupled to a human. A master replica is selected from a segment of the vibrational cardiac data. The master replica is correlated with the segment to obtain a plurality of local maxima. Vibrational cardiac data that were emitted during a diastolic interval are extracted from each heart cycle with the aid of the plurality of local maxima. A two-dimensional space-time frequency power spectrum is processed for Equivalent Rectangular Bandwidth, which provides estimates of the energy produced by turbulent blood flow through a coronary stenosis.

Abstract: A cross arm is provided, for use in a support structure for conductors within an electrical grid. The cross arm is made of weather resistant glass fibre reinforced two component polyurethane based polymers and has a “C” cross sectional shape typically arranged in a double configuration. Pairs of the cross arms are attached to utility poles in a parallel position on opposite sides of the utility pole to meet predefined design conditions. Alternatively, a single cross arm may be attached to a utility pole.

Abstract: A system and method of selling reinsurance includes identifying a reinsurance product and a capacity of the reinsurance product to be sold and calculating a fair risk price for the reinsurance product. The reinsurance product is then offered to potential buyers via an electronic auction, wherein a minimum bid in the electronic auction is determined, at least in part, by the fair risk price. Bids are received from the buyers for portions of the capacity of the reinsurance product and are ranked in accordance with a calculated profitability value and/or time of receipt. Buyers are notified of a status of their respective bids, and particularly, whether bids have been tentatively accepted, partially accepted, or excluded, i.e., rejected. Upon notification of bid status, buyers are given the opportunity to modify bid parameters in an attempt to have their bids ranked at higher a level, thereby increasing the chance of the bid being accepted.

Abstract: A method includes acquiring vibrational cardiac data from an array of N transducers wherein the transducers are coupled to a human. A master replica is selected from a segment of the vibrational cardiac data. The master replica is correlated with the segment to obtain a plurality of local maxima. Vibrational cardiac data that were emitted during a diastolic interval are extracted from each heart cycle with the aid of the plurality of local maxima. A two-dimensional space-time frequency power spectrum is processed for Equivalent Rectangular Bandwidth, which provides estimates of the energy produced by turbulent blood flow through a coronary stenosis.