US 12,271,799 B2
Techniques for disease prediction using machine learning-improved simulations and for generating display elements using simulation results
Audrey Ruple, Blacksburg, VA (US); Johannes Paul Wowra, Darmstadt (DE); John K. Giannuzzi, Wellesley, MA (US); Danna Rabin, New York, NY (US); Christian Debes, Darmstadt (DE); Akash Gupta, Short Hills, NJ (US); Karen Leever, Westport, CT (US); Aliya McCullough, Ardmore, PA (US); and Samantha McKinnon, New York, NY (US)
Assigned to Fetch, Inc., New York, NY (US)
Filed by Fetch, Inc., New York, NY (US)
Filed on Jul. 31, 2023, as Appl. No. 18/362,488.
Application 18/362,488 is a continuation of application No. 18/311,517, filed on May 3, 2023.
Application 18/311,517 is a continuation in part of application No. 17/455,268, filed on Nov. 17, 2021.
Prior Publication US 2023/0376860 A1, Nov. 23, 2023
Int. Cl. G06N 20/20 (2019.01); G06N 20/00 (2019.01); G16H 50/20 (2018.01); G16H 50/50 (2018.01); G16H 70/60 (2018.01)
CPC G06N 20/20 (2019.01) [G06N 20/00 (2019.01); G16H 50/20 (2018.01); G16H 50/50 (2018.01); G16H 70/60 (2018.01)] 24 Claims
OG exemplary drawing
 
1. A method for predictive disease identification, comprising:
applying a plurality of first machine learning models based on training data in order to generate a first set of outputs, wherein the plurality of first machine learning models includes an ensemble of boosting machine learning models and a logistic regression model, wherein the ensemble of boosting machine learning models is sequentially trained using a boosting algorithm in a sequence, wherein misclassifications by a model among the ensemble of boosting machine learning models in the sequence are used to adjust weights of subsequent models among the ensemble of boosting machine learning models in the sequence;
training a second machine learning model based on the first set of outputs, wherein the second machine learning model is a combiner model trained to output a plurality of disease predictor values based on the first set of outputs, wherein each disease predictor value corresponds to a respective disease type of a plurality of disease types, wherein each disease type of the plurality of disease types corresponds to a predetermined group of diseases;
applying the plurality of first machine learning models and the second machine learning model based on features extracted from data including animal characteristics data of an animal, wherein outputs of the plurality of first machine learning models and the second machine learning model include a plurality of disease predictor values, wherein each disease predictor value corresponds to a respective disease type of a plurality of disease types, wherein each disease type of the plurality of disease types corresponds to a predetermined group of diseases;
determining a plurality of simulation parameters based on the plurality of disease predictor values, wherein the plurality of simulation parameters define at least a plurality of time periods for which a plurality of disease contraction simulations are to be run;
running the plurality of disease contraction simulations based on the plurality of simulation parameters in order to obtain simulation results, wherein running the plurality of disease contraction simulations includes providing the plurality of disease predictor values to a simulation engine configured to determine predictions of diseases for animals, wherein the plurality of disease contraction simulations are run for the plurality of time periods defined in the plurality of simulation parameters;
generating disease contraction statistics based on the simulation results; and
determining, based on the disease contraction statistics, at least one disease prediction for the animal.