Parcel growth model training system
An improved parcel growth prediction system that uses parcel data, population data, and artificial intelligence to predict the growth of a geographic area at a micro level (e.g., a real estate parcel level) is described herein. For example, the improved parcel growth prediction system may generate a graph model and apply the graph model as an input to an artificial intelligence model to predict the likelihood that a particular parcel may be developed some time in the future. Ultimately, implementing the improved parcel growth prediction system described herein may lead to more precise placements of infrastructure projects and/or to infrastructure projects that more precisely support the needs of the population of a geographic area as time passes.
1 . A system for training a parcel growth artificial intelligence model, the system comprising:
memory that stores computer-executable instructions; and
a processor in communication with the memory, wherein the computer-executable instructions, when executed by the processor, cause the processor to:
obtain a graph model corresponding to a geographic area, wherein the graph model comprises a first node that corresponds to a first parcel and that is connected to a second node that corresponds to a second parcel;
update a status of the first node to reflect a development status of the first parcel at a first time;
update a status of the second node to reflect a development status of the second parcel at the first time;
train the parcel growth artificial intelligence model using the graph model with the updated status of the first node corresponding to the first time and the updated status of the second node corresponding to the first time;
update a status of the first node to reflect a development status of the first parcel at a second time;
update a status of the second node to reflect a development status of the second parcel at the second time;
apply the graph model with the updated status of the first node corresponding to the second time and the updated status of the second node corresponding to the second time as an input to the trained parcel growth artificial intelligence model to obtain a growth probability; and
validate the trained parcel growth artificial intelligence model based on a comparison of the growth probability with historical data corresponding to a development or lack of development of the first and second parcels by the second time.
2 . The system of claim 1 , wherein the computer-executable instructions, when executed, further cause the processor to:
determine an error rate based on the validation of the trained parcel growth artificial intelligence model; and
re-train the trained parcel growth artificial intelligence model in response to a determination that the error rate exceeds a threshold value.
3 . The system of claim 2 , wherein the computer-executable instructions, when executed, further cause the processor to re-train the trained parcel growth artificial intelligence model using a hyperparameter that is different than a hyperparameter used to train the parcel growth artificial intelligence model.
4 . The system of claim 1 , wherein the development status of the first parcel at the first time comprises one of a developed state, an undeveloped state, or a vacant state.
5 . The system of claim 4 , wherein the undeveloped state is the first parcel has not been developed at least within a threshold time period of the first time.
6 . The system of claim 4 , wherein the developed state is a structure was built on the first parcel at least within a threshold time period of the first time.
7 . The system of claim 1 , wherein the development status of the first parcel at the first time is a developed state, wherein the development status of the second parcel at the first time is a vacant state, wherein the development status of the first parcel at the second time is the developed state, and wherein the development status of the second parcel at the second time is the developed state.
8 . The system of claim 1 , wherein the computer-executable instructions, when executed, further cause the processor to update metadata of the first node to reflect the development status of the first parcel at the first time.
9 . The system of claim 1 , wherein the parcel growth artificial intelligence model comprises a machine learning model.
10 . The system of claim 1 , wherein the second time is after the first time and before a current time.
11 . A computer-implemented method for training a parcel growth artificial intelligence model, the computer-implemented method comprising:
obtaining a graph model corresponding to a geographic area, wherein the graph model comprises a first node that corresponds to a first parcel and that is connected to a second node that corresponds to a second parcel;
updating a status of the first node to reflect a status of the first parcel at a first time;
updating a status of the second node to reflect a status of the second parcel at the first time;
training the parcel growth artificial intelligence model using the graph model with the updated status of the first node corresponding to the first time and the updated status of the second node corresponding to the first time;
updating a status of the first node to reflect a status of the first parcel at a second time;
updating a status of the second node to reflect a status of the second parcel at the second time;
applying the graph model with the updated status of the first node corresponding to the second time and the updated status of the second node corresponding to the second time as an input to the trained parcel growth artificial intelligence model to obtain a growth probability; and
validating the trained parcel growth artificial intelligence model based on a comparison of the growth probability with historical data corresponding to a development or lack of development of the first and second parcels by the second time.
12 . The computer-implemented method of claim 11 , further comprising:
determining an error rate based on the validation of the trained parcel growth artificial intelligence model; and
re-training the trained parcel growth artificial intelligence model in response to a determination that the error rate exceeds a threshold value.
13 . The computer-implemented method of claim 12 , wherein re-training the trained parcel growth artificial intelligence model further comprises re-training the trained parcel growth artificial intelligence model using a hyperparameter that is different than a hyperparameter used to train the parcel growth artificial intelligence model.
14 . The computer-implemented method of claim 11 , wherein the status of the first parcel at the first time comprises one of a developed state, an undeveloped state, or a vacant state.
15 . The computer-implemented method of claim 14 , wherein the developed state is a structure was built on the first parcel at least within a threshold time period of the first time.
16 . The computer-implemented method of claim 11 , wherein updating a status of the first node further comprises updating metadata of the first node to reflect the status of the first parcel at the first time.
17 . A non-transitory, computer-readable medium comprising computer-executable instructions for training a parcel growth artificial intelligence model, wherein the computer-executable instructions, when executed by a computer system, cause the computer system to:
obtain a graph model corresponding to a geographic area, wherein the graph model comprises a first node that corresponds to a first parcel and that is connected to a second node that corresponds to a second parcel;
train the parcel growth artificial intelligence model using the graph model with a status of the first node corresponding to a first time and a status of the second node corresponding to the first time;
apply the graph model with a status of the first node corresponding to a second time and a status of the second node corresponding to the second time as an input to the trained parcel growth artificial intelligence model to obtain a growth probability; and
validate the trained parcel growth artificial intelligence model based on a comparison of the growth probability with historical data corresponding to a development or lack of development of the first and second parcels by the second time.
18 . The non-transitory, computer-readable medium of claim 17 , wherein the computer-executable instructions, when executed, further cause the computer system to:
determine an error rate based on the validation of the trained parcel growth artificial intelligence model; and
re-train the trained parcel growth artificial intelligence model in response to a determination that the error rate exceeds a threshold value.
19 . The non-transitory, computer-readable medium of claim 18 , wherein the computer-executable instructions, when executed, further cause the computer system to re-train the trained parcel growth artificial intelligence model using a hyperparameter that is different than a hyperparameter used to train the parcel growth artificial intelligence model.
20 . The non-transitory, computer-readable medium of claim 17 , wherein the computer-executable instructions, when executed, further cause the computer system to update metadata of the first node to reflect the status of the first parcel at the first time.