WO2022224204A1 - System and method for estimating asset value at a point in time - Google Patents
System and method for estimating asset value at a point in time Download PDFInfo
- Publication number
- WO2022224204A1 WO2022224204A1 PCT/IB2022/053757 IB2022053757W WO2022224204A1 WO 2022224204 A1 WO2022224204 A1 WO 2022224204A1 IB 2022053757 W IB2022053757 W IB 2022053757W WO 2022224204 A1 WO2022224204 A1 WO 2022224204A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- value
- neural network
- asset
- sales
- assets
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000004590 computer program Methods 0.000 claims abstract description 14
- 238000013528 artificial neural network Methods 0.000 claims description 343
- 238000012549 training Methods 0.000 claims description 23
- 238000005315 distribution function Methods 0.000 claims description 14
- 230000000977 initiatory effect Effects 0.000 claims 2
- 230000008859 change Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 238000013459 approach Methods 0.000 description 5
- 230000001419 dependent effect Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000010295 mobile communication Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- RUZYUOTYCVRMRZ-UHFFFAOYSA-N doxazosin Chemical compound C1OC2=CC=CC=C2OC1C(=O)N(CC1)CCN1C1=NC(N)=C(C=C(C(OC)=C2)OC)C2=N1 RUZYUOTYCVRMRZ-UHFFFAOYSA-N 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0206—Price or cost determination based on market factors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0278—Product appraisal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/16—Real estate
Definitions
- the present invention relates to the domain of electronic commerce techniques, and more particular to the field of electronic commerce techniques relating to assets relative to a collection of other assets that are viewed as a basket of goods.
- Figure 1A illustrates a prior art solution for valuing assets.
- sales history data corresponding to the class of assets of interest is obtained for a defined period - e.g. all sales history data more than 'X' months old.
- One solution involves extracting hedonic features from the historical data and using the hedonic features to develop models (e.g. regression models) that can be used to predict sales values for different asset sub-classes in the class of assets.
- the hedonic features based model may be configured to predict or estimate sales values of each of asset types 1 to n based on hedonic data or hedonic features extracted from the historical sales data.
- RSI peerat sales index
- FIG. 1B is a data flow diagram illustrating the solution of Figure 1A - specifically applied to estimates sales values for real estate assets.
- the hedonic feature data models, price median data models and / or RSI models can be used to generate automated valuation models for 1 bedroom properties, 2 bedroom properties, upto 5 + bedroom properties.
- the Repeat Sales Index (RSI) Model generates an index of house prices and relies either on repeat sales of the same property over time, or changes in short term trailing medians - and both methods are subject to limited data coverage errors.
- the existing price median modelling approach means that the resulting price median effectively acts as a basket of goods index but one where the goods being priced varies over time. This leads to a conundrum - a more responsive price median requires the usage of shorter time periods however shortening the time period significantly increases the probability of variations in the assets being used to generate the median and hence will generate price medians that can and often do fluctuate widely.
- the typical response is to set aside the estimate generated by the automated method and instead rely on a professional or expert valuer who is skilled in assessing assets of the type being considered - a practice which by definition will be much more time consuming, expensive, subjective and prone to human errors, bias and potentially prejudice.
- the invention provides methods, systems and computer program products for the estimation of the value of assets based on the intermittent or infrequent sales of other assets with differing characteristics that together act as a basket of goods.
- the methods, systems and computer program products of the present invention may be used for the purposes of valuing real estate assets such as residential, commercial or other real estate properties.
- the invention provides a method for estimating a value for an asset within a class of assets.
- the method comprises implementing at a neural network based modelling platform, the steps of (i) determining a value for said asset based on (a) a reference median value representing a median value of sales prices at time t, if all assets within the class of assets were sold at time t, and (b) a sales delta value representing a predicted difference between the value of the asset and the reference median value.
- the neural network based modelling platform comprises a first neural network and a second neural network.
- the first neural network is configured (i) to receive a first set of input data comprising historical sales data for all assets within the class of assets over a predefined period of time, (ii) to receive one or more sales delta values for one or more assets within the class of assets that has been output from a second neural network, and (iii) to output a value or a probability distribution function representing the reference median value.
- the second neural network is configured (1) to receive from the first neural network a current estimated value for the reference median, (ii) to receive as input, sales delta values corresponding to one or more assets within the class of assets, wherein each said sales delta value comprises a value representing a difference between a historical sales price for such asset and a current value for the reference median that has been output from the first neural network and (iii) to output a sales delta value for said asset; which represents a difference between, or a ratio of, or a percentage value or ratio that has been determined based on, a historical sales price for the asset and a current value for the reference median that has been output from the first neural network
- the value of said asset is determined as:
- AssetPrice actual (t) ReferenceMedtan(t) x ( 1 + SalesDeltaToRM (asset, t) ) wherein:
- AssetPrice actual (t) is an actual sales price recorded for the asset when sold at a point in time
- ReferenceMedian (t) is a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- SalesDeltaToRM (asset; t) is an expected difference value represented in percentage from the Reference Median for the asset at time (t).
- determining the value for said asset is additionally based on an error component representing an error in predicting the sales delta value for the asset
- the error component may be determined based on at least one of first and second discrete sources of error - wherein (i) the first source of error is a first error component representing an error in the sales delta value for the asset and (ii) the second source of error is a second error component representing an error in the reference median value for the asset
- the first error component is determined by a third neural network, wherein the third neural network is configured (i) to receive from the first neural network a current estimated value for the reference median, (11) to receive as input; sales delta values corresponding to one or more assets within the class of assets, wherein each said sales delta value comprises a value representing a difference between a historical sales price for such asset and a current value for the reference median that has been output from the first neural network, and (iii) to output the first error component
- the second error component is determined by a fourth neural network wherein the fourth neural network is configured (i) to receive a first set of input data comprising historical sales data for all assets within the class of assets over a predefined period of time, (11) to receive one or more sales delta values for one or more assets within the class of assets that has been output from the second neural network, and (iii) to output the second error component
- the first neural network may be configured to additionally receive as Input one or more macro or micro economic data inputs, wherein said data inputs determine the output of the sales delta value for said asset
- the neural network based modelling platform Is configured by iteratively training the second neural network using a current Sales Delta value for each iteration.
- Each iteration may comprise (1) providing the current Sales Delta value (SD curr ) to the second neural network as an input value, (ii) evaluating a corresponding output of the second neural network for one or both of accuracy and bias, and (ill) responsive to the output from the second neural network falling to satisfy either an accuracy requirement or a bias threshold (i) determining an estimated Reference Median value (RM est ) based on an output of the current configuration of the second neural network, and (ii) determining an update value for the Reference Median, wherein the update value is determined based on the current value for the Reference Median (Rm curr ) and the estimated Reference Median value (RM est ).
- iterative training of the first neural network may be initiated, starting with a seed value as a current value for a reference median (RM curr1 ), wherein each iteration comprises (i) providing as input to the first neural network, the current value for a reference median (RM curr1 ) on a specific date, (ii) evaluating an output reference median value (RM output ), received as an output from the first neural network for one or both of accuracy and bias, and (ill) responsive to the output reference median value (RM output ) failing to satisfy either an accuracy requirement or a bias threshold, (a) assigning the output reference median value (RM output ) as the current value for a reference median (RM curr1 ), and (b) implementing a successive training iteration upon the first neural network.
- the invention additionally provides a system embodiment comprising a neural network based modelling platform configured for estimating a value for an asset within a class of assets, wherein the value for said asset is based on a reference median value representing a median value of sales prices at time t, if all assets within the class of assets were sold at time t, and a sales delta value representing a predicted difference between the value of the asset and the reference median value.
- the neural network based modelling platform comprises a first neural network and a second neural network.
- the first neural network is configured (1) to receive a first set of input data comprising historical sales data for all assets within the class of assets over a predefined period of time, (ii) to receive one or more sales delta values for one or more assets within the class of assets that has been output from a second neural network, and (iii) to output a value or a probability distribution function representing the reference median value.
- the second neural network may be configured (i) to receive from the first neural network a current estimated value for the reference median, (11) to receive as Input; sales delta values corresponding to one or more assets within the class of assets, wherein each said sales delta value comprises a value representing a difference between a historical sales price for such asset and a current value for the reference median that has been output from the first neural network, and (iii) to output a sales delta value for said asset, which represents a difference between, or a ratio of, or a percentage value or ratio that has been determined based on, a historical sales price for the asset and a current value for the reference median that has been output from the first neural network.
- the neural network based modelling platform may be configured such that the value of said asset is determined as:
- AssetPrice actual (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM (asset; t) ) wherein:
- AssetPrice actual (t) is an actual sales price recorded for the asset when sold at a point in time.
- ReferenceMedian (t) is a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- SalesDeltaToRM (asset t) is an expected difference value represented in percentage from the Reference Median for the asset at time (t).
- determining the value for said asset is additionally based on an error component representing an error in predicting the sales delta value for the asset
- the neural network based modelling platform may be configured to determine the error component based on at least one of first and second discrete sources of error - wherein (i) the first source of error is a first error component representing an error in the sales delta value for the asset; and (ii) the second source of error is a second error component representing an error In the reference median value for the asset
- the first error component is determined by a third neural network, wherein the third neural network is configured (i) to receive from the first neural network a current estimated value for the reference median, (11) to receive as Input; sales delta values corresponding to one or more assets within the class of assets, wherein each said sales delta value comprises a value representing a difference between a historical sales price for such asset and a current value for the reference median that has been output from the first neural network, and (iii) to output the first error component
- the second error component is determined by a fourth neural network, wherein the fourth neural network is configured (1) to receive a first set of Input data comprising historical sales data for all assets within the class of assets over a predefined period of time, (ii) to receive one or more sales delta values for one or more assets within the class of assets that has been output from the second neural network, and (iii) to output the second error component
- the first neural network may be configured to additionally receive as input one or more macro or micro economic data inputs, wherein said data inputs determine the output of the sales delta value for said asset
- the neural network based modelling platform is configured by iteratively training the second neural network using a current Sales Delta value for each iteration.
- Each iteration may comprise (1) providing the current Sales Delta value (SD curr ) to the second neural network as an input value, (ii) evaluating a corresponding output of the second neural network for one or both of accuracy and bias, and (iii) responsive to the output from the second neural network failing to satisfy either an accuracy requirement or a bias threshold (i) determining an estimated Reference Median value (RM est ) based on an output of the current configuration of the second neural network, and (ii) determining an update value for the Reference Median, wherein the update value is determined based on the current value for the Reference Median (Rmcurr)and the estimated Reference Median value (RM est ).
- SD curr current Sales Delta value
- RM est estimated Reference Median value
- each Iteration comprises (1) providing as Input to the first neural network, the current value for a reference median (RM curr1 ) on a specific date, (ii) evaluating an output reference median value (RM output ), received as an output from the first neural network for one or both of accuracy and bias, and (iii) responsive to the output reference median value (RM output ) failing to satisfy either an accuracy requirement or a bias threshold, (a) assigning the output reference median value (RM output ) as the current value for a reference median (RM curr1 ), and (b) implementing a successive training iteration upon the first neural network.
- the Invention also provides a computer program product for estimating a value for an asset within a class of assets.
- the computer program product comprises a non- transitory computer usable medium having computer readable program code embodied therein.
- the computer readable program code comprises instructions for implementing at a neural network based modelling platform, the steps of (i) determining a value for said asset based on (a) a reference median value representing a median value of sales prices at time t; if all assets within the class of assets were sold at time t, and (b) a sales delta value representing a predicted difference between the value of the asset and the reference median value.
- the neural network based modelling platform comprises a first neural network and a second neural network.
- the first neural network is configured (i) to receive a first set of input data comprising historical sales data for all assets within the class of assets over a predefined period of time, (ii) to receive one or more sales delta values for one or more assets within the class of assets that has been output from a second neural network, and (111) to output a value or a probability distribution function representing the reference median value.
- the second neural network is configured (i) to receive from the first neural network a current estimated value for the reference median, (ii) to receive as input; sales delta values corresponding to one or more assets within the class of assets, wherein each said sales delta value comprises a value representing a difference between a historical sales price for such asset and a current value for the reference median that has been output from the first neural network, and (iii) to output a sales delta value for said asset, which represents a difference between, or a ratio of, or a percentage value or ratio that has been determined based on, a historical sales price for the asset and a current value for the reference median that has been output from the first neural network.
- Figure 1A illustrates a prior art solution for valuing assets.
- Figure 1B illustrates the prior art solution for valuing assets as applied to real estate assets.
- Figure 2 illustrates a methodology of the present invention, wherein value of any asset may be represented as a function of a reference median value corresponding to a basket of assets, and a delta value with respect to that median.
- Figure 3 illustrates a system environment comprising an integrated neural network based modelling platform configured for valuing assets in accordance with the teachings of the present invention.
- Figure 4A illustrates a system environment comprising an integrated neural network based modelling platform comprising a first neural network and a second neural network configured for valuing assets in accordance with the teachings of the present invention.
- Figure 4B illustrates a system environment comprising an integrated neural network based modelling platform comprising a first neural network and a second neural network configured for valuing real estate assets in accordance with the teachings of the present invention.
- Figure 5 illustrates a system environment comprising an Integrated neural network based modelling platform comprising a first neural network and a second neural network, and optional third and fourth neural networks configured for valuing assets in accordance with the teachings of the present invention.
- Figure 6 Illustrates a method for training an integrated neural network based modelling platform, in accordance with the teachings of the present invention.
- Figure 7 A illustrates a method implemented by a Reference Median Estimator Neural Network within the integrated neural network based modelling platform, for valuing assets in accordance with the teachings of the present invention.
- Figure 7B illustrates a method implemented by an Error (RM) Estimator Neural Network within the integrated neural network based modelling platform, for estimating the error component representing the error in predicting a Reference Median value for a given asset
- RM Error
- Figure 7C illustrates a method implemented by a Sales Delta Valuation Estimator Neural Network within the integrated neural network based modelling platform, for valuing assets in accordance with the teachings of the present invention.
- Figure 7D illustrates a method implemented by an Error (SD) Estimator Neural Network within the integrated neural network based modelling platform, for estimating an error component representing the error in predicting the sales delta value for a given asset
- SD Error
- Figure 8 illustrates an exemplary computing system configured to implement the present invention.
- the invention provides methods, systems and computer program products for the estimation of the value of assets.
- the methods, systems and computer program products of the present invention may be used for the purposes of valuing real estate assets such as residential, commercial or other real estate properties.
- the invention enables valuation of a specific asset (for example a real estate asset), based on the intermittent or infrequent sales other assets with differing characteristics that together act as a basket of goods.
- the Invention implements a methodology wherein value of any asset may be represented as a function of a reference median value corresponding to a basket of assets, and a delta value with respect to that median.
- the invention implements an asset value determination methodology where the actual asset price or sales prices of an asset may be represented as:
- AssetPrice actual (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM (asset, t) )
- AssetPrice actual (t) Actual sales price recorded for an asset when sold at a point in time
- ReferenceMedian (t) Median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- SalesDeltaToRM Expected delta in % from the Reference Median for a given asset within the target asset class at time (t).
- SalesDeltaToRM may be assigned a value based on determination of AssetPrice actual (t) / ReferenceMedian (t).
- SalesDeltaToRM may be assigned a value based on determination of (AssetPrice actual (t) / ReferenceMedian (t)) - 1.
- the estimated or predicted asset price or sales price of a particular asset may be determined (in accordance with the present invention) as:
- AssetPrice Predicted (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM(asset, t) ) + Error(t)
- AssetPrice Predicted (t) Appropriate price for a specific asset if sold at a point in time
- ReferenceMedian (t) Median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- SalesDeltaToRM Expected delta in % from the Reference Median for a given asset within the target asset class at time (t).
- SalesDeltaToRM may be assigned a value based on determination of AssetPrice actual (t) / ReferenceMedian (t).
- SalesDeltaToRM may be assigned a value based on determination of (AssetPrice actual (t) / ReferenceMedian (t)) - 1
- Error (t) is an error component representing an error in predicting Asset Price.
- the actual property price or sales prices of a house may be represented as:
- PropertyPrice actual (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM (property, t) )
- ReferenceMedian (t) Median of sales prices if all real estate properties within a target class (e.g. within a locality) were sold at time t in a normally functioning market
- SalesDeltaToRM (property, t) may be assigned a value based on determination of PropertyPrice actual (t) / ReferenceMedian (t).
- SalesDeltaToRM (asset, t) may be assigned a value based on a determination of (AssetPrice actual (t) / (ReferenceMedian (t)) - 1.
- the estimated or predicted property price or sales price of a house may be determined (in accordance with the present invention) as:
- the invention enables optimizes the accuracy of an estimated value of an asset or property.
- the invention uses neural networks for (i) implementing the determination of a reference median (ReferenceMedian(t)) for all assets or properties within a target class, (ii) implementing the estimation of a SalesDeltatoRM value (SalesDeltaToRM(property, t) ) for a specific asset or property, and (iii) implementing the estimate of the error value (Error(t) ).
- an integrated neural network based modelling platform 302 receives historical sales data for assets within a target class (for example, for properties within a locality) within a predefined period or range of dates, along with hedonic features associated with individual assets and properties, and optionally macro /micro economic data.
- the Integrated neural network based modelling platform 302 is trained and configured to output (based on the received data) valuations or estimates of individual assets within the target class, optionally along with a forecast standard deviation (FSD) which represents the accuracy or a confidence score associated with the generated valuation or estimate.
- FSD forecast standard deviation
- the integrated neural network based modelling platform 302 preferably comprises at least two discrete or distinct neural networks - one configured as a Reference Median Estimator Neural Network and the other configured as a Sales Delta Valuation Estimator Neural Network. Each of these is described in more detail below.
- Error (t) may be determined, as being composed of at least an error component representing an error in predicting sales delta, i.e. Error salesDelta (asset, t).
- the error component representing an error in predicting sales delta may be estimated by a third neural network within the integrated neural network based modelling platform - i.e. an Error (SD) Estimator Neural Network.
- the Error (SD) Estimator Neural Network may be configured to receive the same inputs as the Sales Delta Valuation Estimator Neural Network.
- the Error(t) as expressed as a percentage may be determined based on two discrete sources of error i.e. a first error component in predicting the Reference Median and a second error component in predicting Sales Delta - i.e.
- Error (t) ( ((100 + Error ReferenceMedlan (t))/100 X (100 + Error salesDelta (asset,t ))/100) - 1) X 100
- the error component representing an error in predicting sales delta may be estimated by a third neural network within the integrated neural network based modelling platform - i.e. an Error (SD) Estimator Neural Network, and the error component representing an error in predicting the Reference Median may be estimated by a fourth neural network within the integrated neural network based modelling platform - i.e. an Error (RM) Estimator Neural Network.
- the Error (SD) Estimator Neural Network may be configured to receive the same inputs as the Sales Delta Valuation Estimator Neural Network
- the Error (RM) Estimator Neural Network may be configured to receive the same inputs as the Reference Median Estimator Neural Network.
- the invention may be implemented through a plurality of discrete neural networks, wherein:
- a first neural network i.e. a Reference Median Estimator Neural Network
- a Reference Median Estimator Neural Network is configured to for estimating the reference median at each instance of time (t) i.e. ReferenceMedian(t)
- a second neural network i.e. a Sales Delta Valuation Estimator Neural Network
- a Sales Delta Valuation Estimator Neural Network is configured for estimating the sales delta value for an asset (or property) i.e. the expected delta in % (or otherwise) from the Reference Median for a given asset (or property) within the target asset class (or target class of properties) at time (t) i.e. SalesDeltaToRM(asset, t) or SalesDeltaToRM (property, t),
- an optional third neural network i.e. an Error (SD) Estimator Neural Network
- SD Error
- Neural Network an optional third neural network that is configured for estimating the error component representing the error in predicting the sales delta value for a given asset (or property) i.e. Error salesDelta (t)
- an optional fourth neural network i.e. an Error (RM) Estimator Neural Network
- RM Error
- Neural Network i.e. an Error (RM) Estimator Neural Network
- the first neural network is configured for generating a stable and accurate value for ReferenceMedian(t).
- the second neural network is configured for generating stable and accurate the sales delta value for each asset under consideration.
- the optional third neural network is configured for estimating the error component representing the error in predicting the sales delta value for a given asset (or property) i.e. ErrorSalesDelta(t).
- the optional fourth neural network is configured for estimating the error component representing the error in predicting the Reference Median value for a given asset (or property) i.e. Error ReferenceMedian (t).
- first and second neural networks, and optionally the third and/or fourth neural networks may be implemented within the integrated neural network based modelling platform 302 for generating automated models for estimating value of assets in accordance with the present invention.
- the accuracy of outputs from the neural networks is optimized.
- certain embodiments of the Invention implement pipelining of the output from each of the first and second neural networks as input to the other of the first and second neural networks, which provides improvements in accuracy of the estimates from each of the two neural networks.
- the second neural network that is configured for generating sales delta values
- the third neural network that is configured for generating the error in predicting the sales delta value for a given asset
- the second neural network that is configured for generating sales delta values
- the third neural network that is configured for generating the error in predicting the sales delta value for a given asset
- the first neural network that is configured for generating values for ReferenceMedian(t)
- the fourth neural network that is configured for generating the error in predicting the Reference Median value for a given asset
- the second neural network that is configured for generating sales delta values
- the third neural network necessarily require to be trained sequentially i.e. the second neural network is trained first, and the third neural network is trained after the second neural network is determined to have achieved acceptable thresholds for accuracy and / or bias.
- the first neural network that is configured for generating values for ReferenceMedian(t)
- the fourth neural network necessarily require to be trained sequentially i.e. the first neural network is trained first, and the fourth neural network is trained after the first neural network is determined to have achieved acceptable thresholds for accuracy and / or bias.
- Figure 4A illustrates a system environment 400 comprising a specific embodiment of an integrated neural network based modelling platform 402 configured in accordance with the teachings of the present invention, wherein integrated neural network based modelling platform 402 is configured for generating automated valuation models for assets in accordance with the teachings of the present invention.
- Figure 4B illustrates a more specific embodiment of system environment 400, wherein the integrated neural network based modelling platform 402 is configured for generating automated valuation models for real estate assets in accordance with the teachings of the present invention.
- the first neural network / Reference Median Estimator Neural Network 4022 receives the following sets of input data: a. the first set of input data comprises historical sales data for all assets within a target class of assets (e.g. all properties within a target locality) over a predefined period of time. b. the sales delta value for one or more assets within the target class of assets (e.g. for one or more properties within the target locality) that has been output from the second neural network / Sales Delta Valuation Estimator Neural Network 4024, and c. optionally, one or more macro /micro economic data inputs including any of unemployment data, working hours data, labor force participation data, wage growth data and / or population growth data.
- the first neural network / Reference Median Estimator Neural Network 4022 Based on these inputs, the first neural network / Reference Median Estimator Neural Network 4022 outputs a value (or a probability distribution function - for example a quantized representation of a probability distribution function - representing a range of values) for the variable ReferenceMedian(t), representing a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- the second neural network / (i.e. a Sales Delta Valuation Estimator Neural Network) 4024 receives the following sets of input data: a. a current estimated value for the variable ReferenceMedian(t) that is output from the first neural network / Reference Median Estimator Neural Network 4022, b. sales delta values corresponding to one or more assets within the target class of assets (e.g.
- parameter data representing one or more attributes of the asset(s) within the target asset class.
- parameter data may represent one or more of relative location of the property within the target locality, side of the street, elevation, relative elevation, heritage / flood plain status, proximity to one or more locations of interest (e.g. proximity to schools, shopping centers, train stations / subways etc.).
- the second neural network / Sales Delta Valuation Estimator Neural Network 4024 outputs a sales delta value (or a probability distribution function - for example a quantized representation of a probability distribution function - representing a range of sales delta values) for each asset or property (i.e. SalesDeltaToRM(asset, t) or SalesDeltaToRM(property, t)), which represents a difference between, or a ratio of, or a percentage value or ratio that has been calculated based on, a historical sales price for a specific asset / property and a current value for the variable ReferenceMedian(t) that has been output from the first neural network / Reference Median Estimator Neural Network 4022.
- a sales delta value or a probability distribution function - for example a quantized representation of a probability distribution function - representing a range of sales delta values
- each asset or property i.e. SalesDeltaToRM(asset, t) or SalesDeltaToRM(property,
- the following values can therefore be determined: a. a value for the variable ReferenceMedian(t), representing a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market, and b. a sales delta value for each asset or property (i.e.
- AssetPrice Predicted (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM(asset, t)) or in a simplified embodiment where SalesDeltaToRM (asset, t) is a direct multiple of the Reference Median value:
- AssetPrice Predicted (t) ReferenceMedian(t) xSalesDeltaToRM (asset, t)) or in an embodiment that factors in error components:
- AssetPrice Predicted (t) ReferenceMedian(t) x ( 1 + SalesDeltaToRM(asset, t)) + Error(t) or in an embodiment that factors in error components and which specifically considers both Of Error SalesDelta (asset, t) and Error ReferenceMedian (t):
- AssetPrice Predtcted (t) ReferenceMedian(t) x (1 + SalesDeltaToRM (asset, t) ) x ( (100 + Error ReferenceMedian (t))/ 100 x (100 + Error SalesDelta (asset; t)) / 100) )
- the first neural network 4022 is dependent on output from the second neural network for determining values of the variable ReferenceMedian(t)
- the second neural network 4024 is dependent on output from the first neural network for determining values of the variable SalesDeltaToRM(asset, t).
- each of the first neural network and the second neural network may each receive "seed values" as inputs from a source that is different from the other neural network (for example by way of user input; or by way of one or more predefined seed values retrieved from a database), for use in place of the outputs from the other of the first and second neural network.
- the invention further optimizes the accurate determination of AssetPrice Predtcted (t) (or PropertyPrice Predtcted (t), where the asset is a real estate property) by avoiding use of a single value estimator for AssetPrice Predtcted (t) (or PropertyPrice Predtcted (t)) or for ReferenceMedian(t), for the purpose of training the neural networks.
- probability density functions respectively representing estimated values for AssetPrice Predtcted (t) (or PropertyPrice Predtcted (t)), and for ReferenceMedian(t) can be generated and used for training the neural networks 4022, 4024.
- Neural networks 4022, 4024 that have been trained using probability density functions, will in turn provide output in the form of an output probability density function that offers multiple insights - for example, a narrow probability density function curve is an indication of a high confidence prediction, whereas a wide probability density function curve is an indication of a low confidence prediction. Additionally, a change in the shape of the curve (wider, double hump etc.) may represent a change in market conditions.
- Figure 5 illustrates a system environment 500 comprising another embodiment of an integrated neural network based modelling platform 502 configured in accordance with the teachings of the present invention, wherein integrated neural network based modelling platform 502 is configured for generating automated valuation models for assets in accordance with the teachings of the present invention.
- the integrated neural network based modelling platform 502 comprises a plurality of discrete neural networks, wherein:
- a first neural network 5022 i.e. a Reference Median Estimator Neural Network
- a Reference Median Estimator Neural Network is configured to for estimating the reference median at each instance of time (t) i.e. ReferenceMedian(t)
- a second neural network 5024 i.e. a Sales Delta Valuation Estimator Neural Network
- a Sales Delta Valuation Estimator Neural Network is configured for estimating the sales delta value for an asset (or property) l.e. the expected delta in % (or otherwise) from the Reference Median for a given asset (or property) within the target asset class (or target class of properties) at time (t) i.e. SalesDeltaToRM(asset, t) or SalesDeltaToRM (property, t),
- a third neural network 5028 i.e. an Error (SD) Estimator Neural Network
- SD Error
- t Error SalesDelta
- a fourth neural network 5026 i.e. an Error (RM) Estimator Neural Network
- RM Error
- t Reference Median
- the first neural network / Reference Median Estimator Neural Network 5022 receives the following sets of input data: a. the first set of input data comprises historical sales data for all assets within a target class of assets (e.g. all properties within a target locality) over a predefined period of time, b. the sales delta value for one or more assets within the target class of assets (e.g. for one or more properties within the target locality) that has been output from the second neural network / Sales Delta Valuation Estimator Neural Network 5024, and c.
- the first set of input data comprises historical sales data for all assets within a target class of assets (e.g. all properties within a target locality) over a predefined period of time
- the sales delta value for one or more assets within the target class of assets e.g. for one or more properties within the target locality
- the first neural network / Reference Median Estimator Neural Network 5022 Based on these inputs, the first neural network / Reference Median Estimator Neural Network 5022 outputs a value (or a probability distribution function - for example a quantized representation of a probability distribution function - representing a range of values) for the variable ReferenceMedian(t), representing a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- the second neural network / Sales Delta Valuation Estimator Neural Network 5024 receives the following sets of input data: a. a current estimated value for the variable ReferenceMedian(t) that is output from the first neural network / Reference Median Estimator Neural Network 5022, b. sales delta values corresponding to one or more assets within the target class of assets (e.g.
- parameter data representing one or more attributes of the asset(s) within the target asset class.
- parameter data may represent one or more of relative location of the property within the target locality, side of the street, elevation, relative elevation, heritage / flood plain status, proximity to one or more locations of interest (e.g. proximity to schools, shopping centers, train stations / subways etc.).
- the third neural network / Error (SD) Estimator Neural Network 5028 receives the following sets of input data: a. a current estimated value for the variable ReferenceMedian(t) that is output from the first neural network / Reference Median Estimator Neural Network 5022, b. sales delta values corresponding to one or more assets within the target class of assets (e.g.
- parameter data representing one or more attributes of the asset(s) within the target asset class.
- parameter data may represent one or more of relative location of the property within the target locality, side of the street; elevation, relative elevation, heritage / flood plain status, proximity to one or more locations of interest (e.g. proximity to schools, shopping centers, train stations / subways etc.).
- the third neural network / Error (SD) Estimator Neural Network 5028 outputs an error component value representing the error in predicting the sales delta value for a given asset (or property) i.e. Error SalesDelta (t).
- the fourth neural network / Reference Median Estimator Neural Network 5026 receives the following sets of input data: a. the first set of input data comprises historical sales data for all assets within a target class of assets (e.g. all properties within a target locality) over a predefined period of time, b. the sales delta value for one or more assets within the target class of assets (e.g. for one or more properties within the target locality) that has been output from the second neural network / Sales Delta Valuation Estimator Neural Network 5024, and c. optionally, one or more macro /micro economic data inputs Including any of unemployment data, working hours data, labor force participation data, wage growth data and / or population growth data.
- the fourth neural network / Error (RM) Estimator Neural Network 5026 outputs an error component value representing the error in predicting the Reference Median value for a given asset (or property) i.e. Error ReferenceMedian (t).
- the following values can therefore be determined: a. a value for the variable ReferenceMedian(t), representing a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market, and b. a sales delta value for each asset or property (i.e.
- AssetPrice Predicted (t) ReferenceMedlan(t) x ( 1 + SalesDeltoToRM(asset, t)) or in a simplified embodiment where SalesDeltoToRM (asset, t) is a direct multiple of the ReferenceMedian value:
- AssetPrice Predicted (t) ReferenceMedian(t) x SalesDeltoToRM (asset, t)) or in an embodiment that factors in error components:
- AssetPrice Predicted (t) ReferenceMedian(t) x ( 1 + SalesDeltoToRM(asset, t)) + Error(t) or in an embodiment that factors in error components and which specifically considers both of Error SalesDelta (asset, t) and ErrorReferenceMedlan(t):
- AssetPrice Predicted (t) ReferenceMedian(t) x (1 + SalesDeltaToRM(asset, t))x (100 + Error ReferenceMedian (t))/ 100 X (100 + Error SalesDelta (asset, t)) / 100
- the first neural network 5022 is dependent on output from the second neural network 5024 for determining values of the variable ReferenceMedian(t)
- the second neural network 5024 is dependent on output from the first neural network for determining values of the variable SalesDeltaToRM(asset, t).
- each of the first neural network and the second neural network may each receive “seed values" as inputs from a source that is different from the other neural network (for example by way of user input; or by way of one or more predefined seed values retrieved from a database), for use in place of the outputs from the other of the first and second neural network.
- Figure 6 illustrates a method for training an integrated neural network based modelling platform 402, 502 in accordance with the teachings of the present invention. In implementing the method of Figure 6:
- Step 602 comprises assigning a seed value as a current value for a Reference Median (RM curr )
- Step 604 comprises determining a current Sales Delta value (SD curr ) based on the current value for the Reference Median (RM curr )
- Step 606 comprises iteratively training the Sales Delta Valuation Estimator N eural Network 4024, 5024 using a current Sales Delta value for each iteration, wherein each iteration comprises (a) providing the current Sales Delta value (SD curr ) to the Sales Delta Valuation Estimator Neural Network 4024, 5024 as an input value, (b) evaluating the output ofthe Sales Delta Valuation Estimator Neural Network 4024, 5024 for one or both of accuracy and bias, and (c) responsive to the output from the Sales Delta Valuation Estimator Neural Network 4024, 5024 failing to satisfy either an accuracy requirement or a bias threshold (1) determining an estimated Reference Median value (RM est ) based on an output of the current configuration of the Sales Delta Valuation Estimator Neural Network 4024, 5024, (ii) optionally processing or smoothing the estimated Reference Median value Rmest using a filter, and (iii) determining an update value for the Reference Median, wherein the update value is determined based on the current value for the Reference Median
- Step 608 is implemented sequentially after step 606, and comprises responding to output from the Sales Delta Valuation Estimator Neural Network 4024, 5024 satisfying either or both of an accuracy requirement and a bias threshold, initiate iterative training of the Reference Median Estimator Neural Network 4022, 5022 starting with a seed value or an initial value as current value for a Reference Median (RM curr1 ), wherein each iteration comprises (a) providing as input to the Reference Median Estimator Neural Network 4022, 5022, the current value for a Reference Median (RM curr1 ) on a specific date, and optionally macro and micro economic data on that data, as input values, (b) evaluating an output Reference Median value (RM output ), received as an output from the Reference Median Estimator Neural Network 4022, 5022 for one or both of accuracy and bias, and (c) responsive to the output Reference Median value (RM output ) failing to satisfy either an accuracy requirement or a bias threshold, (i) assigning the output Reference Median value (RM output ) as the current value
- Figure 7 A illustrates a method implemented by a Reference Median Estimator Neural Network 4022, 5022 within the integrated neural network based modelling platform 402, 502, for valuing assets in accordance with the teachings of the present invention.
- a Reference Median Estimator Neural Network 4022, 5022 within the integrated neural network based modelling platform 402, 502, for valuing assets in accordance with the teachings of the present invention.
- Step 702A comprises receiving at a Reference Median Estimator Neural Network 4022, 5022, inputs comprising (a) a first set of input data comprising historical sales data for all assets within a target class of assets over a predefined period of time, and optionally macro and micro economic data related to the historical sales data (b) the sales delta value for one or more assets within the target class of assets that has been output from a Sales Delta Valuation Estimator Neural Network 4024, 5024, (c) optionally an error component input that has been determined, based on or derived from the error in predicting the sales delta value (i.e.
- Step 704A comprises generating an output from the Reference Median Estimator Neural Network 4022, 5022, wherein the output is determined based on processing of the inputs received at step 702A, said output comprising a value or a probability distribution function for the variable ReferenceMedian(t), representing a median of sales prices if all assets within a target asset class were sold at time t in a normally functioning market
- Figure 7B illustrates a method implemented by an Error (RM) Estimator Neural Network 5026 within the integrated neural network based modelling platform 502, for estimating the error component representing the error in predicting a Reference Median value for a given asset
- RM Error
- Step 702B comprises receiving at an Error(RM) Estimator Neural Network 5026, inputs comprising (a) a first set of input data comprising historical sales data for all assets within a target class of assets over a predefined period of time, and optionally macro and micro economic data related to the historical sales data (b) the sales delta value for one or more assets within the target class of assets that has been output from a Sales Delta Valuation Estimator Neural Network 5024, (c) optionally an error component input that has been determined, based on or derived from the error in predicting the sales delta value (i.e. ErrorSalesDelta) for one or more assets within the target class of assets, that has been output from an Error(SD) Estimator Neural Network 5028
- Step 704B comprises generating an output from the Error (RM) Estimator Neural Network 5026, wherein the output is determined based on processing of the inputs received at step 702B, said output comprising a value for (or that enables determination of) a variable ErrorReferenceMedian(t) representing an error in estimating the Reference Median.
- RM Error
- Step 702C comprises receiving at a Sales Delta Valuation Estimator Neural Network 4024, 5024, inputs comprising (a) a current estimated value for the variable ReferenceMedian(t) that is output from a Reference Median Estimator Neural Network 4022, 5022, (b) sales delta values corresponding to one or more assets within the target class of assets, wherein the sales delta values have been calculated by determining a value representing a difference between, or a ratio of, or a percentage value or ratio that has been calculated based on, a historical sales price for such asset / property and a current value for the variable ReferenceMedian(t) that has been output from the Reference Median Estimator Neural Network 4022, 5022, and (c) optionally an error component input that
- Step 704C comprises generating an output from the Sales Delta Valuation Estimator Neural Network 4024, 5024, wherein the output is determined based on processing of the Inputs received at step 702C, said output comprising a sales delta value or a probability distribution function for each asset or property, which represents the a difference between, or a ratio of, or a percentage value or ratio that has been calculated based on, a historical sales price for a specific asset / property and a current value for the variable ReferenceMedian(t) that has been output from the Reference Median Estimator Neural Network 4022, 5022
- Step 702D comprises receiving at an Error(SD) Estimator Neural Network 5028, inputs comprising (a) a current estimated value for the variable ReferenceMedian(t) that is output from a Reference Median Estimator Neural Network 4022, 5022, (b) sales delta values corresponding to one or more assets within the target class of assets, wherein the sales delta values have been calculated by determining a value representing a difference between, or a ratio of, or a percentage value or ratio that has been calculated based on, a historical sales price for such asset / property and a current value for the variable ReferenceMedian(t) that has been output from the Reference Median Estimator Neural Network 4022, 5022, and (c) optionally an error component input that has been determined,
- Step 704D comprises generating an output from the Error(SD) Estimator Neural Network 5028, wherein the output is determined based on processing of the inputs received at step 702 D, said output comprising a value for (or that enables determination of) the variable Error SalesDelta (asset, t) representing an error in estimating the sales delta value for such asset or property.
- FIG. 8 illustrates an exemplary system 800 for implementing the present invention.
- the illustrated system 800 comprises computer system 802 which in turn comprises one or more processors 804 and at least one memory 806.
- Processor 804 is configured to execute program instructions - and may be a real processor or a virtual processor. It will be understood that computer system 802 does not suggest any limitation as to scope of use or functionality of described embodiments.
- the computer system 802 may include, but is not be limited to, one or more of a general-purpose computer, a programmed microprocessor, a micro-controller, an integrated circuit and other devices or arrangements of devices that are capable of implementing the steps that constitute the method of the present invention.
- Exemplary embodiments of a computer system 802 in accordance with the present invention may include one or more servers, desktops, laptops, tablets, smart phones, mobile phones, mobile communication devices, tablets, phablets and personal digital assistants.
- the memory 806 may store software for implementing various embodiments of the present invention.
- the computer system 802 may have additional components.
- the computer system 802 may include one or more communication channels 808, one or more input devices 810, one or more output devices 812, and storage 814.
- An interconnection mechanism (not shown) such as a bus, controller, or network, interconnects the components of the computer system 802.
- operating system software provides an operating environment for various softwares executing in the computer system 802 using a processor 804, and manages different functionalities of the components of the computer system 802.
- the communication channel(s) 808 allow communication over a communication medium to various other computing entities.
- the communication medium provides information such as program instructions, or other data in a communication media.
- the communication media includes, but is not limited to, wired or wireless methodologies implemented with an electrical, optical, RF, infrared, acoustic, microwave, Bluetooth or other transmission media.
- the Input devlce(s) 810 may Include, but is not limited to, a touch screen, a keyboard, mouse, pen joystick, trackball, a voice device, a scanning device, or any another device that is capable of providing input to the computer system 802.
- the input device(s) 810 may be a sound card or similar device that accepts audio input in analog or digital form.
- the output device(s) 812 may include, but not be limited to, a user interface on CRT, LCD, LED display, or any other display associated with any of servers, desktops, laptops, tablets, smart phones, mobile phones, mobile communication devices, tablets, phablets and personal digital assistants, printer, speaker, CD/DVD writer, or any other device that provides output from the computer system 802.
- the storage 814 may include, but not be limited to, magnetic disks, magnetic tapes, CD-ROMs, CD-RWs, DVDs, any types of computer memory, magnetic stripes, smart cards, printed barcodes or any other transitory or non -transitory medium which can be used to store information and can be accessed by the computer system 802.
- the storage 814 may contain program instructions for implementing any of the described embodiments.
- the computer system 802 is part of a distributed network or a part of a set of available cloud resources.
- the present invention may be implemented in numerous ways including as a system, a method, or a computer program product such as a computer readable storage medium or a computer network wherein programming instructions are communicated from a remote location.
- the present invention may suitably be embodied as a computer program product for use with the computer system 802.
- the method described herein is typically implemented as a computer program product, comprising a set of program instructions that is executed by the computer system 802 or any other similar device.
- the set of program instructions may be a series of computer readable codes stored on a tangible medium, such as a computer readable storage medium (storage 814), for example, diskette, CD-ROM, ROM, flash drives or hard disk, or transmittable to the computer system 802, via a modem or other interface device, over either a tangible medium, including but not limited to optical or analogue communications channel(s) 808.
- the Implementation of the Invention as a computer program product may be in an Intangible form using wireless techniques, including but not limited to microwave, infrared, Bluetooth or other transmission techniques. These instructions can be preloaded into a system or recorded on a storage medium such as a CD-ROM, or made available for downloading over a network such as the Internet or a mobile telephone network.
- the series of computer readable Instructions may embody all or part of the functionality previously described herein.
Abstract
Description
Claims
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2022260821A AU2022260821A1 (en) | 2021-04-23 | 2022-04-22 | System and method for estimating asset value at a point in time |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2021901204 | 2021-04-23 | ||
AU2021901204A AU2021901204A0 (en) | 2021-04-23 | System and method for estimating asset value at a point in time |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022224204A1 true WO2022224204A1 (en) | 2022-10-27 |
Family
ID=83722738
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/IB2022/053757 WO2022224204A1 (en) | 2021-04-23 | 2022-04-22 | System and method for estimating asset value at a point in time |
Country Status (2)
Country | Link |
---|---|
AU (1) | AU2022260821A1 (en) |
WO (1) | WO2022224204A1 (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5361201A (en) * | 1992-10-19 | 1994-11-01 | Hnc, Inc. | Real estate appraisal using predictive modeling |
US7305328B1 (en) * | 2001-12-28 | 2007-12-04 | Fannie Mae | Method and apparatus for predicting and reporting a real estate value based on a weighted average of predicted values |
US7970674B2 (en) * | 2006-02-03 | 2011-06-28 | Zillow, Inc. | Automatically determining a current value for a real estate property, such as a home, that is tailored to input from a human user, such as its owner |
US20140279176A1 (en) * | 2013-03-15 | 2014-09-18 | Redfin Corporation | Provision of real-estate market information |
US20150242747A1 (en) * | 2014-02-26 | 2015-08-27 | Nancy Packes, Inc. | Real estate evaluating platform methods, apparatuses, and media |
US10192275B2 (en) * | 2015-03-30 | 2019-01-29 | Creed Smith | Automated real estate valuation system |
CN109360018A (en) * | 2018-09-27 | 2019-02-19 | 郑州轻工业学院 | A kind of fuzzy zone land price estimation method based on artificial neural network |
CN110837921A (en) * | 2019-10-29 | 2020-02-25 | 西安建筑科技大学 | Real estate price prediction research method based on gradient lifting decision tree mixed model |
US20200402116A1 (en) * | 2019-06-19 | 2020-12-24 | Reali Inc. | System, method, computer program product or platform for efficient real estate value estimation and/or optimization |
-
2022
- 2022-04-22 WO PCT/IB2022/053757 patent/WO2022224204A1/en active Application Filing
- 2022-04-22 AU AU2022260821A patent/AU2022260821A1/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5361201A (en) * | 1992-10-19 | 1994-11-01 | Hnc, Inc. | Real estate appraisal using predictive modeling |
US7305328B1 (en) * | 2001-12-28 | 2007-12-04 | Fannie Mae | Method and apparatus for predicting and reporting a real estate value based on a weighted average of predicted values |
US7970674B2 (en) * | 2006-02-03 | 2011-06-28 | Zillow, Inc. | Automatically determining a current value for a real estate property, such as a home, that is tailored to input from a human user, such as its owner |
US20140279176A1 (en) * | 2013-03-15 | 2014-09-18 | Redfin Corporation | Provision of real-estate market information |
US20150242747A1 (en) * | 2014-02-26 | 2015-08-27 | Nancy Packes, Inc. | Real estate evaluating platform methods, apparatuses, and media |
US10192275B2 (en) * | 2015-03-30 | 2019-01-29 | Creed Smith | Automated real estate valuation system |
CN109360018A (en) * | 2018-09-27 | 2019-02-19 | 郑州轻工业学院 | A kind of fuzzy zone land price estimation method based on artificial neural network |
US20200402116A1 (en) * | 2019-06-19 | 2020-12-24 | Reali Inc. | System, method, computer program product or platform for efficient real estate value estimation and/or optimization |
CN110837921A (en) * | 2019-10-29 | 2020-02-25 | 西安建筑科技大学 | Real estate price prediction research method based on gradient lifting decision tree mixed model |
Also Published As
Publication number | Publication date |
---|---|
AU2022260821A1 (en) | 2023-11-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101936029B1 (en) | Valuation method based on deep-learning and apparatus thereof | |
US8010324B1 (en) | Computer-implemented system and method for storing data analysis models | |
US8180664B2 (en) | Methods and systems for forecasting with model-based PDF estimates | |
CN108399564B (en) | Credit scoring method and device | |
CN108256898B (en) | Product sales prediction method, system and storage medium | |
CN109711860A (en) | Prediction technique and device, storage medium, the computer equipment of user behavior | |
US20200027119A1 (en) | Dynamic promotion analytics | |
CN110555578A (en) | sales prediction method and device | |
CN110704730A (en) | Product data pushing method and system based on big data and computer equipment | |
KR102384860B1 (en) | Operating method of open market platform with improved product upload convenience by providing templates | |
CN107958297B (en) | Product demand prediction method and product demand prediction device | |
CN112668238B (en) | Rainfall processing method, rainfall processing device, rainfall processing equipment and storage medium | |
CN110704706B (en) | Training method and classification method of classification model, related equipment and classification system | |
US20090276290A1 (en) | System and method of optimizing commercial real estate transactions | |
Cobb et al. | Accurate lead time demand modeling and optimal inventory policies in continuous review systems | |
CN114187009A (en) | Feature interpretation method, device, equipment and medium of transaction risk prediction model | |
US20190295196A1 (en) | Comparability score using multi-characteristic dimension reduction method | |
CN113822693A (en) | Method, device, equipment and storage medium for generating user purchasing power evaluation value | |
CN112686690A (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
AU2022260821A1 (en) | System and method for estimating asset value at a point in time | |
CN112287208A (en) | User portrait generation method and device, electronic equipment and storage medium | |
CN113849580A (en) | Subject rating prediction method and device, electronic equipment and storage medium | |
US20210304243A1 (en) | Optimization of markdown schedules for clearance items at physical retail stores | |
EP4258193A1 (en) | Method and apparatus for predicting risk, electronic device, computer readable storage medium | |
JP2019145007A (en) | Project progress prediction device and project progress prediction system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22791237 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: AU2022260821 Country of ref document: AU Ref document number: 2022260821 Country of ref document: AU |
|
ENP | Entry into the national phase |
Ref document number: 2022260821 Country of ref document: AU Date of ref document: 20220422 Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 22791237 Country of ref document: EP Kind code of ref document: A1 |