# US 2023/0066201 A1 · Ensemble machine learning framework for predictive operational load balancing

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

- Number: US 2023/0066201 A1
- Document ID: US20230066201A1
- Status: Published application
- Kind: A1
- Application number: 18/048,556
- Filed: 2022-10-21
- Published: 2023-03-02
- Earliest claimed: 2020-05-13
- Assignees: Optum Services (Ireland) Limited
- Applicants: Optum Services (Ireland) Limited
- Inventors: Kieran O'Donoghue; Michael J. McCarthy; Neill Michael Byrne; David Lewis Frankenfield
- Web page: https://patentgrep.com/patent/US20230066201A1

## CPC

- G06F 9/5083: {Techniques for rebalancing the load in a distributed system}
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G06Q 10/063: Operations research, analysis or management
- G06Q 10/10: Office automation; Time management
- G06Q 50/22: Social work or social welfare, e.g. community support activities or counselling services
- G16H 40/20: for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

There is a need for more effective and efficient constrained-optimization-based operational load balancing. In one example, a method comprises determining constraint-satisfying operator-unit mapping arrangements that satisfy an operator unity constraint and an operator capacity constraint; for each constraint-satisfying operator-unit mapping arrangement, determining an arrangement utility measure; processing each arrangement utility measure using an optimization-based ensemble machine learning model that is configured to determine an optimal operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements; and initiating the performance of one or more operational load balancing operations based on the optimal operator-unit mapping arrangement.

## Claims

### Claim 1

A computer-implemented method for performing predictive operational load balancing for an operational system that is associated with a plurality of operator profiles, a plurality of work unit profiles, and one or more operational condition designations, the computer-implemented method comprising:
determining a plurality of constraint-satisfying operator-unit mapping arrangements that satisfy an operator unity constraint and an operator capacity constraint, wherein: (i) in accordance with the operator unity constraint, each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements maps each work unit profile of the plurality of work unit profiles to an operator profile of the plurality of operator profiles, and (ii) in accordance with the operator capacity constraint, each total operator utilization measure for an operator profile of the plurality of operator profiles under a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements satisfies an estimated capacity of the operator profile;
for each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements, determining an arrangement utility measure that describes an estimated utility of utilizing mappings between the plurality of operator profiles and the plurality of work unit profiles using an optimization-based ensemble machine learning model;
determining an optimization-based ensemble machine learning model based on each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements; and
initiating the performance of one or more operational load balancing operations based at least in part on the optimal operator-unit mapping arrangement.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements is determined based at least in part on: (i) an operator-designation model that describes relationships between the plurality of operator profiles and the one or more operational condition designations, (ii) a unit-designation model that describes relationships between the plurality of work units and the one or more operational condition designations, (iii) an operational condition reward model that describes an estimated reward of each operational condition designation of the one or more operational condition designations, and (iv) an operator cost-capacity model that describes an estimated cost and the estimated capacity for each operator profile of the plurality of operator profiles.

### Claim 3 (depends on 2)

The computer-implemented method of claim 2, wherein:
the operator-designation model describes an operator-designation effectiveness measure for each operator-designation pair of a plurality of operator-designation pairs that comprises an operator profile of the plurality of operator profiles and an operational condition designation of the one or more operational condition designations, and
the operator-designation effectiveness measure for an operator-designation pair of the plurality of operator-designation pairs describes estimated effectiveness of utilizing the operator profile that is associated with the operator-designation pair to addressing the operational condition designation that is associated with the operator-designation pair.

### Claim 4 (depends on 2)

The computer-implemented method of claim 2, wherein:
the unit-designation model describes a unit-designation effectiveness measure for each unit-designation pair of a plurality of unit-designation pairs that comprises a work unit profile of the plurality of work unit profiles and an operational condition designation of the one or more operational condition designations, and
the unit-designation effectiveness measure for a unit-designation pair of the plurality of unit-designation pairs describes estimated effectiveness of performing the work unit profile that is associated with the unit designation pair to addressing the operational condition designation that is associated with the operator-designation pair.

### Claim 5 (depends on 2)

The computer-implemented method of claim 2, wherein determining the arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements comprises:
determining an arrangement reward measure for the constraint-satisfying operator-unit mapping arrangement based at least in part on the unit-designation model, the operator-designation model, and each operational condition reward model for a work unit profile of the plurality of work unit profiles;
determining an arrangement cost measure for the constraint-satisfying operator-unit mapping arrangement based at least in part on each operator cost model for an operator profile of the plurality of operator profiles that is utilized by the constraint-satisfying operator-unit mapping arrangement; and
determining the arrangement utility measure based at least in part on the arrangement reward measure and the arrangement cost measure.

### Claim 6 (depends on 2)

The computer-implemented method of claim 2, wherein:
the optimization-based ensemble machine learning model is a stochastic optimization model, and
each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements describes a stochastic arrangement utility measure for the constraint-satisfying operator-unit mapping arrangement.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, wherein each stochastic arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements is determined based at least in part on at least one of:
an operator-designation probability distribution for each operator-designation probability pair of the plurality of operator-designation pairs that is described by the operator-designation model,
a unit-designation probability distribution for each unit-designation probability pair of the plurality of unit-designation pairs that is described by the unit-designation model,
an operational condition reward distribution for each work unit profile of the plurality of work unit profiles that is described by the operational condition reward model,
an operator cost probability distribution for each operator profile of the plurality of operator profiles that is described by the operator cost-capacity model, and
an operator capacity probability distribution for each operator profile of the plurality of operator profiles that is described by the operator cost-capacity model.

### Claim 8 (depends on 1)

The computer-implemented method of claim 1, wherein performing the one or more operational load balancing operations comprises:
for each operator profile of the plurality of operator profiles, causing presentation of an operator queue user interface that describes an assigned subset of the plurality of work unit profiles that have been mapped to the operator profile according to the optimal operator-unit mapping arrangement.

### Claim 9 (depends on 8)

The computer-implemented method of claim 8, wherein the operator queue user interface presented to an operator profile of the plurality of operator profiles describes a priority score for each work unit profile in the assigned subset.

### Claim 10 (depends on 9)

The computer-implemented method of claim 9, wherein the priority score for a work unit profile in the assigned subset of an operator profile of the plurality of operator profiles is determined based at least in part on at least one of a work unit reward measure for the work unit profile.

### Claim 11 (depends on 1)

The computer-implemented method of claim 1, wherein:
the plurality of work unit profiles describe a plurality of patient profiles,
the plurality of operator profiles describe a plurality of provider profiles, and
the one or more operational condition designations describe one or more medical condition profiles.

### Claim 12 (depends on 1)

The computer-implemented method of claim 1, wherein:
the plurality of work unit profiles describe a plurality of medical record profiles,
the plurality of operator profiles describe a plurality of provider profiles, and
the one or more operational condition designations describe one or more medical condition profiles.

### Claim 13

An apparatus for performing predictive operational load balancing for an operational system that is associated with a plurality of operator profiles, a plurality of work unit profiles, and one or more operational condition designations, the apparatus comprising at least one processor and at least one memory including program code, the at least one memory and the program code configured to, with the processor, cause the apparatus to at least:
determine a plurality of constraint-satisfying operator-unit mapping arrangements that satisfy an operator unity constraint and an operator capacity constraint, wherein: (i) in accordance with the operator unity constraint, each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements maps each work unit profile of the plurality of work unit profiles to an operator profile of the plurality of operator profiles, and (ii) in accordance with the operator capacity constraint, each total operator utilization measure for an operator profile of the plurality of operator profiles under a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements satisfies an estimated capacity of the operator profile;
for each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements, determine an arrangement utility measure that describes an estimated utility of utilizing mappings between the plurality of operator profiles and the plurality of work unit profiles using an optimization-based ensemble machine learning model;
determine an optimal operator-unit mapping arrangement based on each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements; and
initiate the performance of one or more operational load balancing operations based at least in part on the optimal operator-unit mapping arrangement.

### Claim 14 (depends on 13)

The apparatus of claim 13, wherein each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements is determined based at least in part on: (i) an operator-designation model that describes relationships between the plurality of operator profiles and the one or more operational condition designations, (ii) a unit-designation model that describes relationships between the plurality of work units and the one or more operational condition designations, (iii) an operational condition reward model that describes an estimated reward of each operational condition designation of the one or more operational condition designations, and (iv) an operator cost-capacity model that describes an estimated cost and the estimated capacity for each operator profile of the plurality of operator profiles.

### Claim 15 (depends on 14)

The apparatus of claim 14, wherein:
the operator-designation model describes an operator-designation effectiveness measure for each operator-designation pair of a plurality of operator-designation pairs that comprises an operator profile of the plurality of operator profiles and an operational condition designation of the one or more operational condition designations, and
the operator-designation effectiveness measure for an operator-designation pair of the plurality of operator-designation pairs describes estimated effectiveness of utilizing the operator profile that is associated with the operator-designation pair to addressing the operational condition designation that is associated with the operator-designation pair.

### Claim 16 (depends on 14)

The apparatus of claim 14, wherein:
the unit-designation model describes a unit-designation effectiveness measure for each unit-designation pair of a plurality of unit-designation pairs that comprises a work unit profile of the plurality of work unit profiles and an operational condition designation of the one or more operational condition designations, and
the unit-designation effectiveness measure for a unit-designation pair of the plurality of unit-designation pairs describes estimated effectiveness of performing the work unit profile that is associated with the unit designation pair to addressing the operational condition designation that is associated with the operator-designation pair.

### Claim 17

A computer program product for performing predictive operational load balancing for an operational system that is associated with a plurality of operator profiles, a plurality of work unit profiles, and one or more operational condition designations, the computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions configured to:
determine a plurality of constraint-satisfying operator-unit mapping arrangements that satisfy an operator unity constraint and an operator capacity constraint, wherein: (i) in accordance with the operator unity constraint, each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements maps each work unit profile of the plurality of work unit profiles to an operator profile of the plurality of operator profiles, and (ii) in accordance with the operator capacity constraint, each total operator utilization measure for an operator profile of the plurality of operator profiles under a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements satisfies an estimated capacity of the operator profile;
for each constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements, determine an arrangement utility measure that describes an estimated utility of utilizing mappings between the plurality of operator profiles and the plurality of work unit profiles using an optimization-based ensemble machine learning model;
determine an optimal operator-unit mapping arrangement based on each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements; and
initiate the performance of one or more operational load balancing operations based at least in part on the optimal operator-unit mapping arrangement.

### Claim 18 (depends on 17)

The computer program product of claim 17, wherein each arrangement utility measure for a constraint-satisfying operator-unit mapping arrangement of the plurality of constraint-satisfying operator-unit mapping arrangements is determined based at least in part on: (i) an operator-designation model that describes relationships between the plurality of operator profiles and the one or more operational condition designations, (ii) a unit-designation model that describes relationships between the plurality of work units and the one or more operational condition designations, (iii) an operational condition reward model that describes an estimated reward of each operational condition designation of the one or more operational condition designations, and (iv) an operator cost-capacity model that describes an estimated cost and the estimated capacity for each operator profile of the plurality of operator profiles.

### Claim 19 (depends on 18)

The computer program product of claim 18, wherein:
the operator-designation model describes an operator-designation effectiveness measure for each operator-designation pair of a plurality of operator-designation pairs that comprises an operator profile of the plurality of operator profiles and an operational condition designation of the one or more operational condition designations, and
the operator-designation effectiveness measure for an operator-designation pair of the plurality of operator-designation pairs describes estimated effectiveness of utilizing the operator profile that is associated with the operator-designation pair to addressing the operational condition designation that is associated with the operator-designation pair.

### Claim 20 (depends on 18)

The computer program product of claim 18, wherein:
the unit-designation model describes a unit-designation effectiveness measure for each unit-designation pair of a plurality of unit-designation pairs that comprises a work unit profile of the plurality of work unit profiles and an operational condition designation of the one or more operational condition designations, and
the unit-designation effectiveness measure for a unit-designation pair of the plurality of unit-designation pairs describes estimated effectiveness of performing the work unit profile that is associated with the unit designation pair to addressing the operational condition designation that is associated with the operator-designation pair.
