# US 12,001,928 B1 · Systems and methods for artificial-intelligence-assisted prediction generation

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,001,928 B1
- Document ID: US12001928B1
- Status: Granted
- Kind: B1
- Application number: 16/834,271
- Filed: 2020-03-30
- Granted: 2024-06-04
- Earliest claimed: 2019-03-29
- Assignees: Cigna Intellectual Property, Inc.
- Applicants: Cigna Intellectual Property, Inc.
- Inventors: John E. Paton; Armand E. Prieditis
- Web page: https://patentgrep.com/patent/US12001928B1

## CPC

- G06N 20/00: Machine learning
- G06F 17/16: Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)}
- G06F 17/18: for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)}
- G06F 18/24: Classification techniques
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06F 2218/20: by applying autoregressive analysis
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A confidence interval determination system includes a processor and memory. The memory stores a historical data store and instructions. The instructions include, in response to receiving a prediction request, determining a size of each element of the prediction request and obtaining a previous prediction request. The instructions include, in response to the prediction request being different from the previous prediction request, for each element of the prediction request, generating a classifier based on the training data and, for each bin of a set of bins of the element, determining a set of predicted likelihoods using the classifier. The instructions include generating a mean vector prediction and a covariance matrix using the sets of bins generated, determining a predicted outcome and a confidence interval using the mean vector prediction and the covariance matrix, and, in response to the confidence interval exceeding a threshold, outputting the predicted outcome and the confidence interval.

## Claims

### Claim 1

A computerized confidence interval determination system comprising:
at least one processor and
memory coupled to the at least one processor,
wherein the memory stores:
a historical data store including training data relating to a particular category and previous prediction requests and
instructions for execution by the at least one processor and
wherein the instructions include, in response to receiving a prediction request from a user device:
determining a size of each element of the prediction request;
obtaining a previous prediction request from the historical data store;
determining whether a match exists between elements of the prediction request and elements of the previous prediction request, including comparing each element of the prediction request to a respective element of the previous prediction request;
in response to a determination that a match exists, determining whether the training data has changed since the previous prediction request was made;
in response to a determination that the training data has not changed:
obtaining a previously stored classifier, and
using the previously stored classifier to determine a set of predicted likelihoods;
in response to any mismatch between the elements of the prediction request and the elements of the previous prediction request, for each element of the prediction request:
determining a set of bins that correspond to the element based on training data,
generating an updated classifier based on the training data, and
for each bin of the set of bins of the element, determining the set of predicted likelihoods using the updated classifier;
in response to a determination that the training data has changed, for each element of the prediction request:
determining the set of bins that correspond to the element based on training data,
generating the updated classifier based on the training data, and
for each bin of the set of bins of the element, determining the set of predicted likelihoods using the updated classifier;
generating a mean vector prediction and a covariance matrix using the set of bins generated for each element of the prediction request;
determining a predicted outcome and a confidence interval using the mean vector prediction and the covariance matrix; and
in response to a determination that the confidence interval exceeds a threshold, outputting, to the user device, the predicted outcome and the confidence interval.

### Claim 2 (depends on 1)

The system of claim 1 wherein each set of predicted likelihoods includes a probability that an input of the element of the prediction request produces an output categorized in a corresponding bin of the set of bins.

### Claim 3 (depends on 1)

The system of claim 1 wherein the instructions include, for each element of the prediction request, storing the updated classifier.

### Claim 4 (depends on 1)

The system of claim 1 wherein each bin of the set of bins is sized based on the training data.

### Claim 5 (depends on 1)

The system of claim 1 wherein:
the set of bins includes a total number of bins and
each bin of the set of bins is of varying size.

### Claim 6 (depends on 1)

The system of claim 1 wherein the updated classifier for each element is generated based on a value of the training data in each bin.

### Claim 7 (depends on 1)

The system of claim 1 wherein each generated classifier implements a corresponding function.

### Claim 8 (depends on 1)

The system of claim 1 wherein each generated classifier is generated based on at least one of a softmax classifier, a naive Bayes classifier, a probabilistic classifier, and a statistical regression classifier.

### Claim 9 (depends on 1)

The system of claim 1 wherein the particular category is related to at least one of healthcare and weather.

### Claim 10

A confidence interval determination method comprising:
in response to receiving a prediction request from a user device:
determining a size of each element of the prediction request;
obtaining a previous prediction request from a historical data store, wherein the historical data store includes training data relating to a particular category and previous prediction requests;
determining whether a match exists between elements of the prediction request and elements of the previous prediction request including comparing each element of the prediction request to a respective element of the previous prediction request;
in response to a determination that a match exists, determining whether the training data has changed since the previous prediction request was made;
in response to a determination that the training data has not changed:
obtaining a previously stored classifier; and
using the previously stored classifier to determine a set of predicted likelihoods;
in response to any mismatch between the elements of the prediction request and the elements of the previous prediction request, for each element of the prediction request:
determining a set of bins that correspond to the element based on training data,
generating an updated classifier based on the training data, and
for each bin of the set of bins of the element, determining the set of predicted likelihoods using the updated classifier;
in response to a determination that the training data has changed, for each element of the prediction request:
determining the set of bins that correspond to the element based on training data,
generating the updated classifier based on the training data, and
for each bin of the set of bins of the element, determining the set of predicted likelihoods using the updated classifier;
generating a mean vector prediction and a covariance matrix using the set of bins generated for each element of the prediction request;
determining a predicted outcome and a confidence interval using the mean vector prediction and the covariance matrix; and
in response to a determination that the confidence interval exceeds a threshold, outputting, to the user device, the predicted outcome and the confidence interval.

### Claim 11 (depends on 10)

The method of claim 10 wherein each set of predicted likelihoods includes a probability that an input of the element of the prediction request produces an output categorized in a corresponding bin of the set of bins.

### Claim 12 (depends on 10)

The method of claim 10 wherein each generated classifier implements a corresponding function.

### Claim 13 (depends on 1)

The system of claim 1 wherein generating the mean vector prediction and the covariance matrix includes performing calculations based on the set of predicted likelihoods of each bin and one or more bin definitions.

### Claim 14 (depends on 1)

The system of claim 1 wherein determining the set of predicted likelihoods for a first element of the prediction request includes determining, for each bin in the set of bins for the first element, a probability that an input of the first element produces an output categorized in the bin.

### Claim 15 (depends on 1)

The system of claim 1 wherein comparing each element of the prediction request to the respective element of the previous prediction request includes comparing the size of the element to a size of the respective element.

### Claim 16 (depends on 10)

The method of claim 10 wherein generating the mean vector prediction and the covariance matrix includes performing calculations based on the set of predicted likelihoods of each bin and one or more bin definitions.

### Claim 17 (depends on 10)

The method of claim 10 wherein determining the set of predicted likelihoods for a first element of the prediction request includes determining, for each bin in the set of bins for the first element, a probability that an input of the first element produces an output categorized in the bin.

### Claim 18 (depends on 10)

The method of claim 10 wherein comparing each element of the prediction request to the respective element of the previous prediction request includes comparing the size of the element to a size of the respective element.