Granted Patent B2

US 12,561,591 · App. 18/414,211 · Granted Feb 24, 2026

Inventors: William Enerson Harvey (San Francisco, CA); Thomas Janos Atwood (Venice, CA); Marc-Henri Gires (New York, NY)

Assignee: P44, LLC

G06N7/01G06F9/542G06F16/258G06F16/285G06N20/00G06Q10/047

Abstract

Claims (78)

1 . A computer-implemented method of classifying indicators, the method comprising:

accessing, by at least one processor, a sequence of indicators;

detecting, by the at least one processor, at least one duplicate indicator in the sequence of indicators;

removing, by the at least one processor, the at least one duplicate indicator from the sequence of indicators to reduce a computational complexity, the removing resulting in a deduplicated sequence of indicators;

selecting, by the at least one processor, a first indicator in the deduplicated sequence of indicators;

assigning, by the at least one processor, a label to the first indicator;

referencing, by the at least one processor, a milestone graph defining a valid ordered sequence of transitions to milestones and associated constraints;

generating, by the at least one processor, a current search space graph comprising a set of valid paths through a set of milestones in response to the milestone graph and the deduplicated sequence of indicators;

selecting, by the at least one processor, a node on the search space graph in response to the label that was assigned;

defining, by the at least one processor, a state for the node that was selected;

determining, by the at least one processor, that an end of the deduplicated sequence of indicators has been reached; and

returning, by the at least one processor, a classification result comprising a predicted sequence of milestones corresponding to the deduplicated sequence of indicators, to improve an efficiency in finding an optimal solution in a tractable manner.

2 . The computer-implemented method of claim 1 , further comprising:

prior to determining that the end of the sequence of indicators has been reached, determining that the end of the sequence of indicators has not been reached.

3 . The computer-implemented method of claim 2 , further comprising:

after determining that the end of the sequence of indicators has not been reached, selecting a next indicator in the sequence of indicators;

selecting a transition to a next node on the search space graph; and

assigning an additional label to the next indicator in response to the transition that was selected.

4 . The computer-implemented of claim 3 , further comprising:

repeating the selecting the next indicator, the selecting the transition, and the assigning the additional label until the end of the sequence of indicators has been reached.

5 . The computer-implemented method of claim 3 , wherein selecting the transition to the next node on the search space graph comprises:

selecting the transition to the next node on the search space graph in response to satisfying a set of constraints on an edge defining the transition.

6 . The computer-implemented method of claim 1 , wherein selecting the transition to the next node on the search space graph comprises:

selecting the transition to the next node on the search space graph based on a cost function.

7 . The computer-implemented method of claim 1 , further comprising:

forming the sequence of indicators.

8 . A system for classifying indicators, the system comprising:

a memory storing a set of computer-executable instructions; and

at least one processor interfacing with the memory and configured to execute the set of computer-executable instructions to cause the at least one processor to:

access a sequence of indicators,

detect at least one duplicate indicator in the sequence of indicators,

remove the at least one duplicate indicator from the sequence of indicators to reduce a computational complexity, the removing resulting in a deduplicated sequence of indicators,

select a first indicator in the deduplicated sequence of indicators,

assign a label to the first indicator,

reference a milestone graph defining a valid ordered sequence of transitions to milestones and associated constraints,

generate a current search space graph comprising a set of valid paths through a set of milestones in response to the milestone graph and the deduplicated sequence of indicators,

select a node on the search space graph in response to the label that was assigned,

define a state for the node that was selected,

determine that an end of the deduplicated sequence of indicators has been reached, and

return a classification result comprising a predicted sequence of milestones corresponding to the deduplicated sequence of indicators, to improve an efficiency in finding an optimal solution in a tractable manner.

9 . The system of claim 8 , wherein the at least one processor is configured to execute the set of computer-executable instructions to further cause the at least one processor to:

prior to determining that the end of the sequence of indicators has been reached, determine that the end of the sequence of indicators has not been reached.

10 . The system of claim 9 , wherein the at least one processor is configured to execute the set of computer-executable instructions to further cause the at least one processor to:

after determining that the end of the sequence of indicators has not been reached, select a next indicator in the sequence of indicators,

select a transition to a next node on the search space graph, and

assign an additional label to the next indicator in response to the transition that was selected.

11 . The system of claim 10 , wherein the at least one processor is configured to execute the set of computer-executable instructions to further cause the at least one processor to:

repeat the selecting the next indicator, the selecting the transition, and the assigning the additional label until the end of the sequence of indicators has been reached.

12 . The system of claim 8 , wherein to select the transition to the next node on the search space graph, the at least one processor is configured to:

select the transition to the next node on the search space graph in response to satisfying a set of constraints on an edge defining the transition.

13 . The system of claim 8 , wherein to select the transition to the next node on the search space graph, the at least one processor is configured to:

select the transition to the next node on the search space graph based on a cost function.

14 . The system of claim 8 , wherein the at least one processor is configured to execute the set of computer-executable instructions to further cause the at least one processor to:

form the sequence of indicators.

15 . A non-transitory computer readable medium storing a set of computer instructions executable by at least one processor, the set of computer instructions comprising:

instructions for accessing a sequence of indicators;

instructions for detecting at least one duplicate indicator in the sequence of indicators;

instructions for removing the at least one duplicate indicator from the sequence of indicators to reduce a computational complexity, the removing resulting in a deduplicated sequence of indicators;

instructions for selecting a first indicator in the deduplicated sequence of indicators;

instructions for assigning a label to the first indicator;

instructions for referencing a milestone graph defining a valid ordered sequence of transitions to milestones and associated constraints;

instructions for generating a current search space graph comprising a set of valid paths through a set of milestones in response to the milestone graph and the deduplicated sequence of indicators;

instructions for selecting a node on the search space graph in response to the label that was assigned;

instructions for defining a state for the node that was selected;

instructions for determining that an end of the deduplicated sequence of indicators has been reached; and

instructions for returning a classification result comprising a predicted sequence of milestones corresponding to the deduplicated sequence of indicators, to improve an efficiency in finding an optimal solution in a tractable manner.

16 . The non-transitory computer readable medium of claim 15 , wherein set of computer instructions further comprise:

instructions for, prior to determining that the end of the sequence of indicators has been reached, determining that the end of the sequence of indicators has not been reached.

17 . The non-transitory computer readable medium of claim 16 , wherein set of computer instructions further comprise:

instructions for, after determining that the end of the sequence of indicators has not been reached, selecting a next indicator in the sequence of indicators;

instructions for selecting a transition to a next node on the search space graph; and

instructions for assigning an additional label to the next indicator in response to the transition that was selected.

18 . The non-transitory computer readable medium of claim 17 , wherein set of computer instructions further comprise:

instructions for repeating the selecting the next indicator, the selecting the transition, and the assigning the additional label until the end of the sequence of indicators has been reached.

19 . The non-transitory computer readable medium of claim 17 , wherein the instructions for selecting the transition to the next node on the search space graph comprise:

instructions for selecting the transition to the next node on the search space graph in response to satisfying a set of constraints on an edge defining the transition.

20 . The non-transitory computer readable medium of claim 15 , wherein the instructions for selecting the transition to the next node on the search space graph comprise:

instructions for selecting the transition to the next node on the search space graph based on a cost function.

Assignments (5)

Continuity (3)

References Cited (31)