Patent Issued for Apparatus and method for resource allocation prediction and modeling, and resource acquisition offer generation, adjustment and approval (USPTO 11954623): Assurant Inc.
2024 APR 26 (NewsRx) -- By a
The assignee for this patent, patent number 11954623, is
Reporters obtained the following quote from the background information supplied by the inventors: “Many of today’s network environments are dynamically resource-constrained, at least in the sense that the need for resources, and the nature of the needed resources, can change rapidly and significantly over time and geography. Some of the technical challenges that hinder the effective and efficient allocation of resources in such environments are compounded in situations where the supply, utility, and/or value of the needed resources changes over time. Additionally, in this regard, acquisition of resources for a particular time and/or geography can change significantly. Technical challenges in data compilation, analysis, visualization, and manipulation associated with conventional systems hinder efficient resource acquisition planning. The inventors of the invention disclosed herein have identified these and other technical challenges, and developed the solutions described and otherwise referenced herein.”
In addition to obtaining background information on this patent, NewsRx editors also obtained the inventors’ summary information for this patent: “An apparatus, computer program product, and method are therefore provided in accordance with an example embodiment in order permit the efficient determining of one or more channels and/or related conditions through which a particular resource set may be effectively distributed. In this regard, the method, apparatus and computer program product of an example embodiment provide for the creation of predicted channel and condition data set that can be stored within a renderable object and otherwise presented to a user via an interface of a client device.
“Moreover, the method, apparatus, and computer program product of an example embodiment provide for use of the machine learning model in connection with the determination and retrieval of a predicted channel and condition data set determined based at least in part on context data associated with a particular resource set to be distributed at a time in the future.
“In an example embodiment, an apparatus is provided, the apparatus comprising a processor and a memory, the memory comprising instructions that configure the apparatus to: receive a request data object from a client device associated with a user; extract, from the message request data object, a request data set, wherein the request data set is associated with a first set of resources; receive a first context data object, wherein the first context data object is associated with one or more resource distribution channels; retrieve a predicted channel and condition data set, wherein retrieving the predicted channel and condition data set comprises applying the request data set and the first context data object to a first model; and generate a control signal causing a renderable object comprising the predicted channel and condition data set to be displayed on a user interface of the client device associated with the user.
“In another example embodiment, a computer program product is provided, the computer program product comprising at least one non-transitory computer-readable storage medium having computer-executable program code instructions stored therein, the computer-executable program code instructions comprising program code instructions configured to: receive a request data object from a client device associated with a user; extract, from the message request data object, a request data set, wherein the request data set is associated with a first set of resources; receive a first context data object, wherein the first context data object is associated with one or more resource distribution channels; retrieve a predicted channel and condition data set, wherein retrieving the predicted channel and condition data set comprises applying the request data set and the first context data object to a first model; and generate a control signal causing a renderable object comprising the predicted channel and condition data set to be displayed on a user interface of the client device associated with the user.
“In another example embodiment, a method for determining a predicted future demand for resources in a dynamic environment is provided, the method comprising: receiving a request data object from a client device associated with a user; extracting, from the message request data object, a request data set, wherein the request data set is associated with a first set of resources; receiving a first context data object, wherein the first context data object is associated with one or more resource distribution channels; retrieving a predicted channel and condition data set, wherein retrieving the predicted channel and condition data set comprises applying the request data set and the first context data object to a first model; and generating a control signal causing a renderable object comprising the predicted channel and condition data set to be displayed on a user interface of the client device associated with the user.”
The claims supplied by the inventors are:
“1. A method for allocating a constrained resources set in a dynamic environment, the method comprising: receiving, from a plurality of client devices, at least a first request data object associated with a first channel profile and a second request data object associated with a second channel profile, wherein the first request data object is associated with at least a first future resource volume and a first future time interval and the second request data object is associated with at least a second future resource volume and a second future time interval; assigning the first channel profile and the second channel profile to a first tier from amongst a plurality of tiers; determining, based on the assigned first tier and the first request data object, that the first channel profile satisfies each of a plurality of threshold conditions; determining, based on the assigned first tier and the second request data object, that the second channel profile satisfies each of the plurality of threshold conditions; generating channel context data synthesized from a plurality of non-uniform data sets associated with dynamic resource demand and wherein the plurality of non-uniform data sets is acquired from a plurality of external systems by: scraping a first non-uniform data set from a first external system via a network, retrieving and extracting a second non-uniform data set from a second external system comprising an external social media site; retrieving a stored third non-uniform data set from a repository; periodically updating the third non-uniform data set via data received at regular intervals from a third external system; receiving a fourth non-uniform data set from a fourth external system, the fourth non-uniform data set comprising external environment data; wherein each of the first non-uniform data set, the second non-uniform data set, the third non-uniform data set, and the fourth non-uniform data set from the plurality of external systems comprises occluded relevant components to be utilized with an allocation model comprising a machine learning model; determining at least a first third-party identifier corresponding to at least a portion of the first non-uniform data set; determining at least a second third-party identifier corresponding to at least a portion of the second non-uniform data set; determining at least a third third-party identifier corresponding to at least a portion of the third non-uniform data set; determining at least a fourth third-party identifier corresponding to at least a portion of the fourth non-uniform data set; wherein at least two of the first third-party identifier, the second third-party identifier, the third third-party identifier, or the fourth third-party identifier are not identical; determining first attribute types corresponding to a plurality of first resource attribute values of the first third-party identifier; determining second attribute types corresponding to a plurality of second resource attribute values of the second third-party identifier; determining third attribute types corresponding to a plurality of third resource attribute values of the third third-party identifier; determining fourth attribute types corresponding to a plurality of fourth resource attribute values of the fourth third-party identifier; identifying standardized resource attribute values corresponding to each standardized resource identifier of a plurality of standardized resource identifiers, wherein the standardized resource attribute values associated with the standardized resource identifiers; generating mapping scores for the first third-party identifier, the second third-party identifier, the third third-party identifier, and the fourth third-party identifier relative to one of the standardized resource identifiers by comparing the first resource attribute values, the second resource attribute values, the third resource attribute values, and the fourth resource attribute values with the standardized resource attribute values; determining a that the mapping scores exceed a threshold value indicating that the first third-party resource identifier, the second third-party resource identifier, the third third-party identifier, and the fourth third-part identifier correspond to the one standardized resource identifier; normalizing, scaling, and combining at at least the portion of data in the first non-uniform data set, the portion of data in the second non-uniform data set, the portion of data in the third non-uniform data set and the portion of data in the fourth non-uniform data set in a synergized data set linked to the standardized resource identifier, wherein the channel context data comprises the synergized data set; and validating the synergized data set for machine learning using a data sufficiency model, wherein validating the synergized data set comprise testing the data relative to an accuracy threshold of the data sufficiency model, wherein the synergized data set exceeds the accuracy threshold; training the machine learning model of an allocation model using at least the synergized data set of the channel context data using a supervised learning process, the supervised learning process comprising receiving a plurality of user confirmations from a user interface of a client device during the training to improve an output of the model; in response to determining that the first channel profile and the second channel profile satisfy each of the plurality of threshold conditions and in real time or near real time relative to receipt of the first request data object and the second request data object, applying a first resource request set associated with the first request data object, a second resource request set associated with the second request data object, a constrained plurality of resources, and the first assigned tier to the allocation model by: applying the constrained plurality of resources, the first resource request set, and the second resource request set to the machine learning model; wherein the machine learning model comprises a multivariate adaptive regression splines (MARS) model; wherein applying the first resource request set and the second resource request set to the MARS model comprises utilizing a hinge function to automatically select variables associated with the first resource request set and the second resource request set and values associated with the variables for knots of the binge function to model an interaction between two or more of the variables to generate a decay set for a predetermined time window; applying the decay set for the predetermined time window to a logistic regression model to determine a predicted resource allocation set comprising predicted resource allocations at least associated with the first channel profile and the second channel profile for the constrained plurality of resources; and outputting the predicted resource allocation set; in real time or near real time relative to receipt of the first request data object and the second request data object and based on the predicted resource allocation set, generating at least one control signal that: dynamically allocates a first plurality of resources from the constrained plurality of resources and causes transmission of the first plurality of resources to a first system associated with the first channel profile; and dynamically allocates a second plurality of resources from the constrained plurality of resources and causes transmission of the second plurality of resources to a second system associated with the second channel profile, wherein the first plurality of resources and the second plurality of resources are determined by the predicted resource allocations in the predicted resource allocation set; wherein the at least one control signal is transmitted during each of the predetermined time window, the first future time interval, and the second future time interval; wherein the at least one control signal comprises renderable data transmitted to graphical user interfaces associated with each of the first system and the second system, updating the training of the machine learning model to generate an updated allocation model with new training data derived from the predicted resource allocation set; subsequent to updating the training of the machine learning model, receiving at least a third request data object associated with a third channel profile, wherein the third request data object is associated with at least a third future resource volume and a third future time interval; generating, via the updated allocation model, a subsequent predicted resource allocation set; and in real time or near real time relative to receipt of the third request data object, generating at least one control signal that dynamically allocates a third plurality of resources from a second constrained plurality of resources and causes transmission of the third plurality of resources to a third system associated with the third channel profile.
“2. The method of claim 1, further comprising applying a plurality of tiering parameters and a first request parameter from the first request data object to a first machine learning model to generate first predicted resource volume data associated with the first channel profile, wherein the plurality of tiering parameters comprises an entropy parameter associated with a channel profile.
“3. The method of claim 2, further comprising scaling the entropy parameter associated with the first channel profile based at least in part on assigning the entropy parameter associated with the first channel profile to a position in a ranked list of entropy parameters.”
There are additional claims. Please visit full patent to read further.
For more information, see this patent: Caltabiano, Brett. Apparatus and method for resource allocation prediction and modeling, and resource acquisition offer generation, adjustment and approval.
(Our reports deliver fact-based news of research and discoveries from around the world.)
University of Johannesburg Reports Findings in Health Economics (A Hierarchical Bayesian approach to small area estimation of health insurance coverage in Ethiopian administrative zones for better policies and programs): Economics – Health Economics
Researchers Submit Patent Application, “Systems And Methods For Generating, Maintaining, And Using An Asset Record On A Blockchain”, for Approval (USPTO 20240119532): Patent Application
Advisor News
Annuity News
Health/Employee Benefits News
Life Insurance News