Abstract: A system and method for Medical Claims Risk Score (MCRS) algorithmic underwriting includes a predictive machine learning model configured to generate underwriting decisions on electronic applications. MCRS underwriting applies word embedding modeling, such as GloVe (global vectors), to transform high dimensional MC records into single-code word vectors. These single-code word vectors are employed in regression modeling, and may include summarized embedding coordinates aggregated at the applicant level. Regression modeling uses medical claim codes data and underwriting decision data stored for historical underwriting applicants to train a random forest model to predict relative mortality risk for underwriting applicants. A risk rating may be derived from the underwriting decision data based upon standard quantitative risk ratings of a plurality of risk classes. Other inputs to the random forest model may include cohort level applicant profile data, such as applicant issue age and sex.

Abstract: Disclosed herein are embodiments of systems, methods, and products comprises an analytic server, which determines user health attributes by tracking the user's behaviors and activities within a predetermined space. The analytic server receives tracking data from a set of sensors installed in the predetermined space. The sensors track a beacon worn by the user. The analytic server determines micro-locations and user behaviors based on the tracking data. The analytic server determines the coordinates of the sensors based on the sensor identifiers and maps the coordinates to regions by referring to a floor plan map. The analytic server determines the user behaviors and activities by aggregating the micro-locations and regions the user visited at different time. The analytic server determines the user's health score based on the micro-locations and user behaviors by executing an artificial intelligence model. The analytic server determines a recommendation of premium based on the health score.

Abstract: The present disclosure relates to a method for operating a lighting device for a motor vehicle, wherein the lighting device has at least an exterior lamp and a control device. The exterior lamp has an illuminating area having at least two independently controllable segments each comprising at least one light source, wherein, in at least one dynamic operating mode of the control device, for at least one dynamic group of the segments comprising at least two segments a pseudo-randomized or randomized, temporally variable lighting operation of the segments of the dynamic group takes place.

Abstract: A system and method for dynamic model training of a predictive machine learning model accesses data points of a training dataset including a plurality of model covariates. The predictive machine learning model is configured to generate an output including a risk rank representative of a mortality risk. The method selects one of the covariates and generates a historical data distribution for the selected covariate by applying the model to the training dataset including a plurality of historical application records. The method determines a current data distribution for the selected covariate. When comparison of the current data distribution with the historical data distribution indicates a data distribution shift exceeding a predetermined threshold, the method automatically updates parameters of the predictive machine learning model and retrains the predictive machine learning model using the updated parameters.

Abstract: Discussed herein are methods and systems for an interdependent series/suite of AI models. In one embodiment, a processor receives projection assumption inputs from a user device and executes a population builder machine learning model to predict a dynamic adjustment table. It applies the model to population data to generate a value population file, which simulates a subset of the population based on the predicted table. The file contains value cells representing instances of a product. The device then runs a mortality machine learning model to determine mortality data for the product using the simulated population. Finally, it executes a flow projection model to generate a projection report for the product, incorporating mortality data and projection assumptions.

Abstract: A suite of fluidless predictive machine learning models includes a fluidless mortality module, smoking propensity model, and prescription fills model. The fluidless machine learning models are trained against a corpus of historical underwriting applications of a sponsoring enterprise, including clinical data of historical applicants. A data appended procedure supplements historical applications data with public records and credit risks. Various features of this data are engineered for improved predictive characteristics. Fluidless models are trained by application of a random forest ensemble including survival, regression and classification models. The trained models produce high-resolution, individual mortality scores. A fluidless underwriting protocol runs these predictive models to assess mortality risk and other risk attributes of a fluidless application that excludes clinical data to determine whether to present an accelerated underwriting offer.

Abstract: Embodiments disclosed herein disclose a back-end computer to generate a risk score and a front-end visualization engine to hierarchically display the generated risk core. The back-end computer users a machine learning model for a stepwise perturbation from a digital reference profile until a user profile to be score is reached. The computer may calculate intermediate risk score for each perturbation and calculate the final risk score after all the perturbations are completed. The front-end visualization engine generates an interactive hierarchical display showing information associated with the risk score calculation. More specifically, the visualization engine may show a filtered list of users sharing one or more attributes with the user profile, a visual rendering of the top factors contributing to the risk score, and individual input values within a factor; and juxtapose the scores and attributes of the user profile in the graphical information display of the associated population.

Abstract: An underwriting estimator predictive machine learning model receives as inputs a limited number of details about an applicant, and outputs an immediate underwriting estimate of risk class. A preliminary pre-screening review redirects applicants with one or more screening impairments to a human-in-the-loop quick quote process. Model inputs include estimator inputs data that are pre-selected from the dataset of impairments data after excluding the screening impairments from the dataset of impairments. The underwriting estimator model may incorporate alternative pathways that output individualized underwriting estimates for some applicants and cohort-level marginal distributions for other applicants. Model outputs also include explanation files providing interpretability of underwriting estimates. The explanation files may include additive feature attribution data and rule based natural language explanations.

Abstract: The present disclosure relates to a method for operating a lighting device for a motor vehicle, wherein the lighting device has at least an exterior lamp and a control device. The exterior lamp has an illuminating area having at least two independently controllable segments each comprising at least one light source, wherein, in at least one dynamic operating mode of the control device, for at least one dynamic group of the segments comprising at least two segments a pseudo-randomized or randomized, temporally variable lighting operation of the segments of the dynamic group takes place.

Abstract: Disclosed herein are embodiments of systems, methods, and products comprises an analytic server, which determines user health attributes by tracking the user’s behaviors and activities within a predetermined space. The analytic server receives tracking data from a set of sensors installed in the predetermined space. The sensors track a beacon worn by the user. The analytic server determines micro-locations and user behaviors based on the tracking data. The analytic server determines the coordinates of the sensors based on the sensor identifiers and maps the coordinates to regions by referring to a floor plan map. The analytic server determines the user behaviors and activities by aggregating the micro-locations and regions the user visited at different time. The analytic server determines the user’s health score based on the micro-locations and user behaviors by executing an artificial intelligence model. The analytic server determines a recommendation of premium based on the health score.

Abstract: A suite of fluidless predictive machine learning models includes a fluidless mortality module, smoking propensity model, and prescription fills model. The fluidless machine learning models are trained against a corpus of historical underwriting applications of a sponsoring enterprise, including clinical data of historical applicants. Fluidless models are trained by application of a random forest ensemble including survival, regression, and classification models. The trained models produce high-resolution, individual mortality scores. A fluidless underwriting protocol runs these predictive models to assess mortality risk and other risk attributes of a fluidless application that excludes clinical data to determine whether to present an accelerated underwriting offer.

Abstract: A suite of fluidless predictive machine learning models includes a fluidless mortality module, smoking propensity model, and prescription fills model. The fluidless machine learning models are trained against a corpus of historical underwriting applications of a sponsoring enterprise, including clinical data of historical applicants. A data appended procedure supplements historical applications data with public records and credit risks. Various features of this data are engineered for improved predictive characteristics. Fluidless models are trained by application of a random forest ensemble including survival, regression and classification models. The trained models produce high-resolution, individual mortality scores. A fluidless underwriting protocol runs these predictive models to assess mortality risk and other risk attributes of a fluidless application that excludes clinical data to determine whether to present an accelerated underwriting offer.

Abstract: Embodiments disclosed herein disclose a back-end computer to generate a risk score and a front-end visualization engine to hierarchically display the generated risk core. The back-end computer users a machine learning model for a stepwise perturbation from a digital reference profile until a user profile to be score is reached. The computer may calculate intermediate risk score for each perturbation and calculate the final risk score after all the perturbations are completed. The front-end visualization engine generates an interactive hierarchical display showing information associated with the risk score calculation. More specifically, the visualization engine may show a filtered list of users sharing one or more attributes with the user profile, a visual rendering of the top factors contributing to the risk score, and individual input values within a factor; and juxtapose the scores and attributes of the user profile in the graphical information display of the associated population.

Abstract: The invention concerns a device for quantifying the dexterity of the fingers of a hand, comprising a main body, several systems for measuring the movement and/or the force applied by a finger in a pressure direction, each system for measuring comprising a sensor for deformation, the sensor for deformation comprising a deformable body, characterized in that each system for measuring also comprises a guide bearing, integral with the main body, a shaft having a bearing surface, the bearing surface being in contact with the deformable body, a tube, adapted to slide in translation in the guide bearing and around the shaft in the pressure direction, having a head adapted to fix the finger to the tube, a first spring connecting the bearing to the tube, a second spring connecting the shaft to the tube, the first spring and the second spring being pre-tensioned, so as to pre-tension the deformable body.

Abstract: Embodiments disclosed herein disclose a back-end computer to generate a risk score and a front-end visualization engine to hierarchically display the generated risk core. The back-end computer users a machine learning model for a stepwise perturbation from a digital reference profile until a user profile to be score is reached. The computer may calculate intermediate risk score for each perturbation and calculate the final risk score after all the perturbations are completed. The front-end visualization engine generates an interactive hierarchical display showing information associated with the risk score calculation. More specifically, the visualization engine may show a filtered list of users sharing one or more attributes with the user profile, a visual rendering of the top factors contributing to the risk score, and individual input values within a factor; and juxtapose the scores and attributes of the user profile in the graphical information display of the associated population.

Abstract: The present invention relates to a new method for quantifying key components of manual dexterity. The present invention also provides methods for diagnosing impaired upper limb and/or hand function in patients based on how these components are affected.