US-20260127030-A1 - COMPUTER-IMPLEMENTED METHODS AND SYSTEMS FOR OPTIMIZING RESOURCE ALLOCATION

US20260127030A1US 20260127030 A1US20260127030 A1US 20260127030A1US-20260127030-A1

Abstract

Computer systems and methods for determining a difference (e.g., decrease or increase) from a resource allocation in an initial proposal for a recipient to form an updated proposal based on the recipient's sensitivity to the difference, wherein the recipient's sensitivity to the difference is determined from a sensitivity model which maps a set of one or more attributes of the recipient to one of a plurality of sensitivity levels. A difference (e.g., discount or increase) from the resource allocation in the initial proposal is then determined from the identified sensitivity level. The sensitivity model may be generated by statistical analysis of historical records for the resource.

Inventors

Anna Keshabyan-Prasad
Sarah Kahlon Sinha
Sheldon Lu-Ming TUNG
Dhiren Kakkar
Yu Ling Ip
SNEHA DESAI
JEAN-CHRISTOPHE BOUËTTÉ

Assignees

THE TORONTO-DOMINION BANK

Dates

Publication Date: 20260507
Application Date: 20241101

Claims (20)

1 . A system for determining a difference from a resource allocation in an initial proposal for a recipient to form an updated proposal for the recipient, the system comprising: a memory, a communication interface, and at least one processor operatively coupled to the memory and the communication interface; the at least one processor configured to: generate a sensitivity model by: obtaining a set of one or more recipient attributes, obtaining a plurality of historical records, each historical record comprising information indicating: a difference between a resource allocation in an initial proposal and an updated proposal for a historical recipient, whether the resource was provided to the historical recipient, and settings for the set of one or more recipient attributes, and sub-dividing the plurality of historical records into a plurality of final segments using statistical analysis such that historical records in a same final segment have settings for the set of one or more recipient attributes that are correlated with a similar level of sensitivity to the difference between the resource allocation in the initial proposal and the updated proposal; and select a difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model.
2 . The system of claim 1 , wherein selecting the difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model comprises: receiving, for the recipient, a setting for each recipient attribute in the set of one or more recipient attributes; mapping, based on the settings for the set of one or more recipient attributes, the recipient to a final segment; and identifying the difference from the resource allocation in the initial proposal for the recipient based on the final segment mapped to the recipient.
3 . The system of claim 2 , wherein generating the sensitivity model further comprises associating each final segment with a difference from the resource allocation in the initial proposal; and identifying the difference from the resource allocation for the recipient based on the final segment mapped to the recipient comprises identifying the difference from the resource allocation in the initial proposal associated with the final segment mapped to the recipient as the difference from the resource allocation in the initial proposal for the recipient.
4 . The system of claim 1 , wherein each final segment of the plurality of final segments is associated with a different sensitivity tier of a plurality of sensitivity tiers.
5 . The system of claim 1 , wherein sub-dividing the plurality of historical records into the plurality of final segments using statistical analysis comprises: sub-dividing the plurality of historical records into a plurality of segments, wherein all the historical records in a same segment of the plurality of segments have a same set of settings for the set of one or more recipient attributes; performing a correlation analysis on the historical records within each segment to identify a relationship between (i) the difference between the resource allocation in the initial proposal and the updated proposal and (ii) providing the resource to the historical recipient; assigning each segment a correlation-based sensitivity rating based on the correlation analysis; dividing the segments into a plurality of correlation-based groups based on the correlation-based sensitivity ratings; and combining the segments in each correlation-based group of the plurality of correlation-based groups to form the final segments.
6 . The system of claim 5 , wherein performing the correlation analysis on the historical records in a segment comprises calculating a correlation coefficient based on the historical records in the segment, wherein the correlation coefficient represents a relationship between (i) the difference between the resource allocation in the initial proposal and the updated proposal, and (ii) and providing the resource to the historical recipient.
7 . The system of claim 6 , wherein the correlation coefficient is a Pearson correlation coefficient.
8 . The system of claim 6 , wherein the correlation-based sensitivity rating for each segment is equal to the correlation coefficient for that segment.
9 . The system of claim 5 , wherein dividing the segments into the plurality of correlation-based groups based on the correlation-based sensitivity ratings comprises dividing the segments into quintiles based on the correlation-based sensitivity ratings.
10 . The system of claim 5 , wherein sub-dividing the plurality of historical records into the plurality of final segments using statistical analysis further comprises, prior to performing the correlation analysis: assigning each segment a discretion-based sensitivity rating based on the differences between the resource allocation in the initial proposal and the updated proposal in the historical records in that segment; dividing the segments with less than a predetermined number of historical records into a plurality of discretion-based groups based on the discretion-based sensitivity rating; and combining the segments in each discretion-based group to form a new segment.
11 . The system of claim 10 , wherein the discretion-based sensitivity rating for each segment is an average difference between the resource allocation in the initial proposal and the updated proposal in the historical records in that segment.
12 . The system of claim 1 , wherein the resource is a financial resource, and the initial proposal allocates an initial price or rate for the financial product.
13 . The system of claim 1 , wherein the resource is a mortgage for a property, the initial proposal allocates an initial rate for the mortgage, and the difference from the resource allocation in the initial proposal and the updated proposal is a discount from the initial rate for the mortgage.
14 . The system of claim 13 , wherein the set of one or more recipient attributes comprises one or more of: whether or not the property is a rental property, a number of days before a closing of the property, whether the recipient is a builder, a construction type of the property, a term type of the mortgage, and an amount of the mortgage.
15 . The system of claim 1 , wherein the set of one or more recipient attributes comprise X recipient attributes that are most statistically significant to sensitivity and X is an integer greater than or equal to one.
16 . A method for determining a difference from a resource allocation in an initial proposal to form an updated proposal, the method executed in a computing environment comprising one or more processors, a communication interface, and memory, and the method comprising: generating a sensitivity model by: obtaining a set of one or more recipient attributes, obtaining a plurality of historical records, each historical record comprising information indicating: a difference between a resource allocation in an initial proposal and an updated proposal for a historical recipient, whether the resource was provided to the historical recipient, and settings for the set of one or more recipient attributes, and sub-dividing the plurality of historical records into a plurality of final segments using statistical analysis such that historical records in a same final segment have settings for the set of one or more recipient attributes that are correlated with a similar level of sensitivity to the difference between the resource allocation in the initial and updated proposals; and selecting a difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model.
17 . The method of claim 16 , wherein selecting the difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model comprises: receiving, for the recipient, a setting for each recipient attribute in the set of one or more recipient attributes; mapping, based on the settings for the set of one or more recipient attributes for the recipient, the recipient to a final segment; and identifying the difference from the resource allocation in the initial proposal for the recipient based on the final segment mapped to the recipient.
18 . The method of claim 16 , wherein sub-dividing the plurality of historical records into the plurality of final segments using statistical analysis comprises: sub-dividing the plurality of historical records into a plurality of segments, wherein all the historical records in a same segment of the plurality of segments have a same set of settings for the set of one or more recipient attributes; performing a correlation analysis on the historical records within each segment to identify a relationship between (i) the difference between the resource allocation in the initial proposal and the updated proposal, and (ii) providing the resource to the historical recipient; assigning each segment a correlation-based sensitivity rating based on the correlation analysis; dividing the segments into a plurality of correlation-based groups based on the correlation-based sensitivity ratings; and combining the segments in each correlation-based group of the plurality of correlation-based groups to form the final segments.
19 . The method of claim 18 , wherein sub-dividing the plurality of historical records into the plurality of final segments using statistical analysis further comprises, prior to performing the correlation analysis: assigning each segment a discretion-based sensitivity rating based on the differences between the resource allocation in the initial proposal and the updated proposal in the historical records in that segment; dividing the segments with less than a predetermined number of historical records into a plurality of discretion-based groups based on the discretion-based sensitivity rating; and combining the segments in each discretion-based group to form a new segment.
20 . A non-transitory computer readable medium storing computer executable instructions which, when executed by at least one computer processor, cause the at least one computer processor to carry out a method of determining a difference from a resource allocation in an initial proposal for a recipient to form an updated proposal, the method comprising: generating a sensitivity model by: obtaining a set of one or more recipient attributes, obtaining a plurality of historical records, each historical record comprising information indicating: a difference between a resource allocation in an initial proposal and an updated proposal for a historical recipient, whether the resource was provided to the historical recipient, and settings for the set of one or more recipient attributes, and sub-dividing the plurality of historical records into a plurality of final segments using statistical analysis such that historical records in a same final segment have settings for the set of one or more recipient attributes that are correlated with a similar level of sensitivity to the difference between the resource allocation in the initial and updated proposals; and selecting a difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model.

Description

TECHNICAL FIELD The disclosed example embodiments relate to computer-implemented methods and systems for determining a recommended difference (e.g., decrease or increase) from an initial proposal for resource allocation. BACKGROUND There are many situations where a first entity is in charge of allocating resources to a second entity. In some cases, the first entity may make an initial proposal for resource allocation to the second entity; and if the second entity is not satisfied with the initial proposal, the first entity may make an updated proposal. The resource that is allocated may be any suitable resource such as, but not limited to, computing resources. For example, in a computing system an operating system may initially propose allocating a process a certain, e.g., default, number of computing resources (e.g., a certain amount of memory and/or a certain amount of central processing unit (CPU) time). The process may not be satisfied with the default number of resources that it has been allocated and thus may request an updated proposal. The operating system may then provide an updated proposal that allocates an increased number of resources (e.g., an increased amount of memory and/or increased CPU time) to the process. The operating system may want to optimize the trade-off between (i) allocating the process sufficient resources so that the process will accept the proposal and (ii) not allocating too many resources so that the operating system can support other processes. It may be desirable in such situations to be able to automatically select an allocation of resources for the updated proposal that balances such competing interests. SUMMARY The following summary is intended to introduce the reader to various aspects of the detailed description, but not to define or delimit any invention. A first aspect provides a system for determining a difference from a resource allocation in an initial proposal for a recipient to form an updated proposal for the recipient, the system comprising: a memory, a communication interface, and at least one processor operatively coupled to the memory and the communication interface; the at least one processor configured to: generate a sensitivity model by: obtaining a set of one or more recipient attributes, obtaining a plurality of historical records, each historical record comprising information indicating: a difference between a resource allocation in an initial proposal and an updated proposal for a historical recipient, whether the resource was provided to the historical recipient, and settings for the set of one or more recipient attributes, and sub-dividing the plurality of historical records into a plurality of final segments using statistical analysis such that historical records in a same final segment have settings for the one or more recipient attributes that are correlated with a similar level of sensitivity to the difference between the resource allocation in the initial proposal and the updated proposal; and select a difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model. Selecting the difference from the resource allocation in the initial proposal for the recipient based on the sensitivity model may comprise: receiving, for the recipient, a setting for each recipient attribute in the set of one or more recipient attributes; mapping, based on the settings for the set of one or more recipient attributes, the recipient to a final segment; and identifying the difference from the resource allocation in the initial proposal for the recipient based on the final segment mapped to the recipient. Generating the sensitivity model may further comprise associating each final segment with a difference from the resource allocation in the initial proposal; and identifying the difference from the resource allocation for the recipient based on the final segment mapped to the recipient may comprise identifying the difference from the resource allocation in the initial proposal associated with the final segment mapped to the recipient as the difference from the resource allocation in the initial proposal for the recipient. Each final segment of the plurality of final segments may be associated with a different sensitivity tier of a plurality of sensitivity tiers. Sub-dividing the plurality of historical records into the plurality of final segments using statistical analysis may comprise: sub-dividing the plurality of historical records into a plurality of segments, wherein all the historical records in a same segment of the plurality of segments have a same set of settings for the set of one or more recipient attributes; performing a correlation analysis on the historical records within each segment to identify a relationship between (i) the difference between the resource allocation in the initial proposal and the updated proposal and (ii) providing the resource to the historical recipient; assigning each segment a c