JP2003526848A - System and method for evaluating a loan portfolio using fuzzy clustering - Google Patents

Abstract

A system and method for evaluating a loan portfolio using clustering logic. The asset acquisition logic includes acquiring a plurality of assets and attaching a plurality of variables to the plurality of assets in an asset database. The segmentation logic examines multiple assets and variables using a model. The fuzzy clustering logic calculates the value of a plurality of assets and variables, and the profit analysis logic calculates the profit of the plurality of assets. The method evaluates the loan portfolio using clustering logic. The method includes acquiring a plurality of assets and attaching a plurality of variables to the assets. The plurality of assets and variables are inspected using a model, and the values of the plurality of assets and variables are calculated using fuzzy clustering. Thereby, the profits of the plurality of assets are calculated.

Description

Detailed Description of the Invention

REFERENCE TO RELATED APPLICATIONS This application is hereby incorporated by reference in its entirety, in 1999.
US Provisional Application No. 60 / 168,499 filed on Dec. 2, 2014, titled “system and Method for Valuing Comm
erial Loan Using Fuzzy Clustering
and Kanolwdge Engineering for GECS C
commercial finance (systems and methods for evaluating commercial loans using fuzzy clustering and knowledge engineering for GECS commercial finance).

[0002]

BACKGROUND OF THE INVENTION

This disclosure relates to valuing loan portfolios, and more particularly, describes systems and methods for valuing loan portfolios using clustering logic.

In general, the valuation process involves determining the value of income / cash flow generating assets. This is a relatively simple process applicable to the projected flow of returns from securities, stocks, leaseholds, wells and loans. The valuation process is performed to determine the asset value at a given time. The time value of financial concepts and the risk / return concept are key elements in this valuation process. Simply stated, the valuation of an asset is equal to the current net value of estimated future profits, which is commonly measured in terms of cash flow.

Currently, it is difficult to estimate the profit valuation of an asset portfolio. In general,
Securities underwriters must examine each loan in the portfolio one at a time. This is very tedious and time consuming.

[0005]

[Outline of the Invention]

This disclosure describes systems and methods for valuing loan portfolios using clustering logic. Briefly, in the architecture, the system is implemented as follows. The asset acquisition logic acquires multiple assets and attaches multiple variables to multiple assets in the asset database. Segmentation logic looks at multiple assets and variables using a model. The fuzzy clustering logic calculates the values of multiple assets and variables, and the profit analysis logic calculates the profits of multiple assets.

The present disclosure can also be viewed as providing a method for evaluating a loan portfolio using clustering logic. In this respect, the method
The following steps can be broadly outlined. (1) Acquire a plurality of assets and attach a plurality of variables to the assets (2) Examine the plurality of assets and variables using a model (3) To calculate the values of the plurality of assets and variables (4) Calculate profit of multiple assets using fuzzy clustering. These assets are
These include (but are not limited to) loan portfolios, securities, stocks and leaseholds.

[0007]

DETAILED DESCRIPTION OF THE INVENTION

Shown in FIG. 1 is a high-level block diagram of asset valuation using a fuzzy clustering process. As shown and described in detail below, this process
Includes acquiring numerous portfolios of loans 2A-2Z. The portfolio of these loans is then reconstructed using the Portfolio Reconstruction Process 3. Next, the reconstructed portfolio is supplied to the integrated loan information process 4, which retrieves the loan information from the loan information database 5 and integrates the loan information into the reconstructed portfolio. Thereafter, the reconstructed portfolio having the attached loan information 5 is output as the reconstructed portfolios 6A to 6H. Then, these reconstructed portfolios 6A ~
6H is supplied to the (expanded) HELTR process 7 of the present invention. The HELTR process 7 provides a method for forming an asset valuation using the fuzzy clustering of the present invention. Valuation of an asset using the fuzzy clustering method of the present invention retrieves the above loan data and then provides an estimated cash flow and risk assessment for each of the reconstructed portfolios of assets 8.

As shown in FIG. 2, computer system 21 generally includes an operating system 32 as well as a processor 22 and memory 31 (eg, RAM, ROM, hard disk, CD-ROM, etc.). The processor 22 uses the memory 3 via the local interface 23 (eg bus).
Receive code and data from 1. Instructions from the user can be sent by using input devices such as (but not limited to) mouse 24 and keyboard 25. This action input and the resulting output is displayed on the display terminal 26. Asset valuation using the fuzzy clustering system 50 can use the modem or network card 27 to access other computers and resources on the network.

FIG. 2 also shows a data acquisition process 60, a variable selection process 80, and a hierarchical segmentation process 100 existing in the memory area 31.
And fuzzy clustering process 120 and underwriting review process 1
Asset evaluation using a fuzzy clustering system 50 including 80 and 80. The database 33 is also shown as resident in the memory area 31. These components are described in more detail in Figures 2-12. Memory area 31 may be, for example (but not limited to) electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, equipment, devices or propagation media. More specific examples of memory area 31 (non-exhaustive list) include electrical connections (electronic) with one or more wires, movable computer disks (magnetic), random access, etc.
Memory (RAM) (magnetic), read-only memory (ROM) (magnetic), erasable programmable read-only memory (EPROM) or flash memory)
(Magnetic), optical fiber (optical), and movable small disk read-only memory (
CDROM) (optical).

Shown in FIG. 3 is an example flow of a system and method for asset valuation using the fuzzy clustering system 50 of the present disclosure. The following description of the asset valuation system and method using the fuzzy clustering system 50 uses the example loan portfolio. Assets may include (but are not limited to) loans, insurance policies, securities, stocks, leasehold properties and other property.

The asset valuation system and method using the fuzzy clustering system 50 includes a data acquisition process 60 that includes acquiring data. For example, in this disclosure, the loan portfolio and the debtor payment history data set,
Loan background information, including credit analysis data sets, car loan and mortgage loan data sets, and industry-specific data sets are used. The data acquisition process 60 uses divide-and-conquer to handle large amounts of asset data. The asset data is input to the variable selection process 80. The variable selection process 80 identifies the critical loan variables for credit reviews, or the variables that have the most outstanding ability and separate the various loan groups.

Both the loan portfolio data collected in the data acquisition process 60 and the critical variables identified in the variable selection process 80 are input to the hierarchical segmentation process 100. Hierarchical segmentation process 100 divides an entire portfolio of assets (ie, loans in this example) into a number of bins based on a predetermined critical variable selected by those considering the credibility of the portfolio. Hierarchical segmentation
After the process 100 is performed, the split assets (ie loans) are further classified by the fuzzy clustering process 120. The fuzzy clustering process 120 classifies each of the divided bins into a predetermined number of clusters based on the natural structure of the asset data. After this classification has been performed by the fuzzy clustering process 120, the classification of assets (ie loans) is further processed by an underwriting review process 180 that assigns estimated cash flows and risk ratings to each asset in the cluster. . The estimated cash flow and risk assessment for each asset in the cluster is
Output for use in portfolio credibility analysis. Data acquisition process 60, variable selection process 80, hierarchical segmentation process 1
00, fuzzy clustering process 120 and underwriting review process 180 are described in more detail below with respect to FIGS.

Shown in FIG. 4 is a flow chart of an example of an asset valuation using the fuzzy clustering system 50. First, an asset valuation using the fuzzy clustering system 50 performs a data acquisition process 60 at step 51. The data acquisition process 60 will be described in more detail with reference to FIG.

Next, the asset valuation using the fuzzy clustering system 50 is step 5.
At 2, the variable selection process 80 is performed. The variable selection process 80 uses the asset data captured in the data acquisition process 60. The variable selection process 80 is described in more detail with reference to FIG.

Asset valuation using fuzzy clustering system 50 then performs a hierarchical segmentation process 100. The hierarchical segmentation process 100 divides a portfolio of assets (ie, loans) into a number of bins preset by the user based on the critical variables they identify. Hierarchical segmentation process 120 is described in more detail with reference to FIG.

Asset valuation using fuzzy clustering system 50 then performs a fuzzy clustering process at step 54. Further, the fuzzy clustering process 120 divides each of the bins identified in the hierarchical segmentation process 100 into a user preset number of clusters based on the natural structure of the asset data. The fuzzy clustering process 120 is described in more detail with reference to FIG.

Asset valuation using the fuzzy clustering system 50 performs an underwriting review process 180 at step 55. Underwriting review process 1
80 assigns an estimated cash flow and risk rating to each of the clusters identified by the fuzzy clustering process 120. The underwriting review process 180 is described in more detail with reference to FIG. At step 59, the asset valuation using the fuzzy clustering system 50 ends.

Shown in FIG. 5 is a block diagram illustrating an exemplary type of asset database used in performing attachment of asset-related variables to an asset portfolio. The data acquisition process 60 includes the step of acquiring asset related data. This process is
In general, it involves the attachment of data related to the assets in the asset portfolio. This disclosure illustrates these concepts using the example loans described throughout this disclosure.
Using this example, the loan is integrated with a relevant data reference to the asset information. Preferably, loan asset data in loan asset database 60 is referenced and integrated into multiple databases.

For example, as shown in FIG. 5, the loan asset database of this example includes, but is not limited to, debtor payment history data set 35, public credit analysis data set 36, personal credit analysis data set. It includes records from a variety of different universal files or databases, including set 37, auto loan and mortgage loan data set data 38, and industry specific data set 39. Preferably, the loan asset records in the loan asset database 60 are integrated with the data referenced above. This is a hierarchical segmentation process 100 and fuzzy
It is useful in identifying the critical variables used during the execution of the clustering process 120. Data scrubbing is performed on the collected data before pulling useful information from the composite database. For example, data scrubbing includes, but is not limited to, detecting outliners, creating or deleting missing values, deriving relevant variables from raw data, and so on.

Shown in FIG. 6 is a table of one implementation example of the variable selection process 80. In the variable selection process 80, the user identifies variables that are considered critical. In this example, the variable selection process 80 has identified 11 variables used by the fuzzy clustering process 120. As shown in the figure, what is associated with each of the variables that are considered critical is the encoding scheme that represents that variable, along with the associated classification and range of values for that variable.

Shown in FIG. 7 is an example of a hierarchical segmentation model created by the hierarchical segmentation process 100 for an example loan portfolio. One example of a hierarchical segmentation model applied by the hierarchical segmentation process 100 is CART. CART is a well-known statistical algorithm for recursive trees and is used for hierarchical segmentation. The idea behind this regression tree is to divide the loan portfolio into a given number of classifications, with respect to the critical variables preset by the user:
The goal is to make each classification homogeneous.

Shown in FIG. 7 is the result of applying the hierarchical segmentation process of the CART model. The resulting regression tree has an example loan portfolio split using three critical variables. These three critical variables for this example include the loan instrument, the loan type, and the last payment for the loan. The resulting regression tree uses these three critical variables to partition the loan portfolio into 6 bins. These partitions can be represented as a tree structure using the CART model. Once the loan portfolio is split using the critical variables, the fuzzy clustering process 120 can perform a fine-grained partitioning of each of these predetermined bins.

Shown in FIG. 8 is the fuzzy clustering process 12 of the present invention.
2 is a flow chart of one example implementation of 0. The example implementation shown in FIG. 8 uses an example of a split loan portfolio created by the hierarchical segmentation process 100.

The fuzzy clustering process 120 first begins at step 121. Next, at step 122, the fuzzy clustering process computes a fuzzy clustering average by performing FCM calculation process 140. This FCM calculation process 140 will be described in more detail with reference to FIG. In step 123, the fuzzy clustering process 120 calculates intra-cluster and inter-cluster variance by box plot. Calculation of intra-cluster and inter-cluster variance with this box plot is a diagnostic check on the final result. This diagnostic check is done by examining the corresponding box plots within and between clusters, respectively. Boxes for intra-cluster and inter-cluster distribution
The plot is described in more detail with reference to Figure 11 (A and B).

At step 124, the fuzzy clustering process 120 determines whether the intra-cluster and inter-cluster variances are compact enough or only one cluster remains. Determining if the clustering is compact enough is solved by determining whether the intra-cluster variance is minimized and the inter-cluster variance is maximized by the FCM calculation process performed at step 122. If at step 124 it is determined that the clustering is compact enough or only one remains, then the fuzzy clustering process 120 ends at step 139.

If at step 124 it is determined that the clustering is not compact enough, the fuzzy clustering process 120 proceeds at step 125 to
Acquire the first pair of the next set of cluster centroids. The fuzzy clustering process 120 calculates the distance between each pair of cluster centroids and stores this distance in a distance matrix at step 126. At step 131, the fuzzy clustering process 120 determines if the clustering centroid pair under test still exists. If the cluster-centroid pair under examination still exists, the fuzzy clustering process 120 returns and repeats steps 125-131.

When there are no more cluster centroid pairs to process, the fuzzy clustering process 120 uses the distance matrix to create a dendrogram for each pair of cluster centroids at step 132. Fuzzy clustering process 120 examines this dendrogram in step 133 to identify possible integrations of cluster centroids. At step 133, the process examines the dendrogram to identify possible integrations of centroid clusters. Example dendrograms are described in more detail with reference to FIGS. 11A-11D.

At step 134, the fuzzy clustering process 120 determines which cluster centroids can be integrated. If there is a possible integration of cluster centroid pairs, the fuzzy clusterlong process 120 returns and repeats steps 122-124. If the fuzzy clustering process 120 determines that there is no possible cluster centroid integration, the fuzzy clustering process 120 ends at step 139. This process is described in more detail with reference to Figures 11A and 11B.

Shown in FIG. 9 is an FCM calculation process 140. First, in step 141, the number of clusters and the weighting index are input to the FCM calculation process 140. The FCM calculation process 140 then obtains a first cluster at step 142. At step 143, the first (next) data point is obtained. The FCM calculation process 140 then randomizes the degree of membership of each point in each cluster in step 144. The membership degree μ _ik is defined by:

[0030]

[Equation 1]

[0031] As is apparent, as the X _k is close to V _i, degree of membership μ _ik of data points X _k of cluster St. in Lloyd V _i increases. At the same time, μ _ik decreases as X _k moves away from V _j (other clusters). At step 145, the FCM calculation determines if all data points are present in the current cluster randomized at step 145. If in step 145 it is determined that all data points are not randomized, the FCM calculation process 140 returns and repeats steps 143-145.

If the FCM calculation process 140 determines that all data points for the current cluster have been randomized, then the FCM calculation process 140 determines at step 146 that all data points are available. Determine if it was randomized to the cluster. If FCM calculation process 140 determines that not all clusters have their data points randomized, then FCM calculation process 140 returns and repeats steps 142-146.

If the FCM calculation process determines that all data points in all clusters have been randomized, then FCM calculation process 140 calculates centroids for all data points in step 147. To do. The i-th cluster centroid V _i is defined by the following equation.

[0034]

[Equation 2]

As can be seen, V _i is the cluster centroid, a weighted sum of the coordinates of X _k . Where k is the number of data points.

A desired number of clusters c and each cluster centroid V _i (i = 1, 2,
.., c), the FCM calculation process 140 converges to a solution for V _i that represents either the local minimum or saddle point of the cost function. As with most non-linear optimization problems, the solution quality of the FCM calculation process 140 depends heavily on the choice of the initial value, the number c and the initial cluster centroid V _i .

Next, at step 148, the FCM calculation process 140 calculates an objective function. This objective function is defined by the following equation.

[0038]

[Equation 3]

[0039] Where n is the number of data points, c is the number of clusters, and Xk is the kth data point.
Data point, Vi is the i-th cluster centroid, μ_ikIs the i-th
The membership degree of the kth data in the raster, m is a constant greater than 1 (typical
Therefore, m = 2). Note that μ_ikIs a real number and is in the range [0, 1]. μ _ik = 1 means that the i-th data is absolutely in the k-th cluster
, Μ_ik= 0 means that the i-th data is absolutely not in the k-th cluster.
To taste. μ_ik= 0.5, the i-th data is partially k-th up to 0.5 degree
Means in a cluster of. Clearly, each data point is unique
Exactly belong to a cluster and have partial membership to other clusters
If not present, the cost function is minimized. That is, each data point
There is no ambiguity in assigning to the cluster to which it belongs.

At step 149, the FCM calculation process 140 determines if the values of the objective function calculation converge. If the objective function calculation does not converge, the FCM calculation process 140
Proceeds to steps 151-155. If, in step 149, the FCM calculation process 140 determines that the value of the objective function converges, the FCM calculation process proceeds to step 1
It ends at 59.

At step 151, CM calculation process 140 obtains a first cluster.
At step 152, a first data point is obtained. Then, the process 153
At, the FCM process 140 updates the membership degree for each point in each cluster. At step 154, FCM calculation process 140 determines if all data points in the current cluster have been updated. F if not all data points in the current cluster have been updated
CM calculation process 140 returns and repeats steps 152-154. If all data points in the current cluster have been updated, then in step 155 the FCM calculation process 140 determines if each data point in all clusters has been updated.

If, at step 155, FCM calculation process 140 determines that all data points have been updated for all clusters, FCM calculation process 14
0 returns and repeats steps 151-155. If, in step 155, FCM calculation process 140 determines that all data points in all clusters have been updated, then FCM calculation process 140 returns to step 147, as described above.
Repeat ~ 149.

Shown in FIG. 10 is an underwriting review process 180. The underwriting review process 180 is performed after the entire portfolio has been split by the fuzzy clustering process 120. Underwriting review process 18
In 0, each cluster is reviewed and assigned a composite score called HELTR. HELTR is H: high cash flow (high cas)
h flow), E: Expected cash flow (expected cash f)
low), L: Low cash flow, T: cash flow timing month, R: borrower risk assessments
Sent of Borrower). In essence, the HELTR score captures the estimated cash and range of cash, the timing of cash flows, and the risk associated with each cluster.

First, the underwriting review process 180 begins at step 181. At step 182, the first loan segment is obtained and becomes the current loan segment. At step 183, the first cluster in the current loan segment is obtained. At step 184, the underwriting review process 180 calculates the cash flow score and cash flow timing for the current cluster in the current loan segment. At step 185, the underwriting review process 180 calculates a risk rating for purchases of all clusters in the current loan segment. At step 186, the underwriting review process determines if an assessment of the current cluster in the current loan segment has been performed. If there are still clusters in the current loan segment, the underwriting review process 180 returns and repeats steps 183-186.

At step 187, the underwriting review process 180 determines if all clusters have been reviewed in all loan segments. If underwriting review process 180 determines that not all clusters in all loan segments have been reviewed, then underwriting review process 180 returns and repeats steps 182-187. Underwriting review process 180
If it is determined that all clusters in all loan segments have been reviewed, the underwriting review process 180 ends at step 189.

Illustrated in FIGS. 11A and 11B is 6 created in step 123 (FIG. 8).
Intra-cluster and inter-cluster dispersion for an example of one cluster. As shown in FIG. 11B, the average distance from all data points to the centroid data point is 1.0, and as shown in FIG. 11A, the average distance of the data points in cluster 1 is 0. It is 6. This shows that clustering is very compact. Therefore, the intra-cluster variance is minimized and the inter-cluster variance is maximized by the FCM calculation process 140.

Shown in FIGS. 11C and 11D are a tree diagram of 20 clusters and a tree diagram of 6 clusters, respectively. It should be noted that each dendrogram starts with the data points forming individual clusters, and data points or clusters that are close to each other have been successfully integrated. Typically, it takes 2-3 iterations of the fuzzy clustering process 120 to get the optimal amount of clusters. As shown in FIG. 11C, 12 of the 20 centroids need to be integrated. Therefore, in the above example, fuzzy clustering
The process 120 is re-executed with 6 clusters with the results shown in FIG. 11D.

Shown in FIG. 12 is the HEL for the six clusters shown in FIG. 11D.
It is an example of TR. As shown, the HELTR table contains data for each centroid identified in the example of FIG. 7 and data for each cluster within each identified centroid. The data used in this HELTR table includes: whether the loan is mortgage guaranteed, installment credit, requires notice, is the final payment made, is the loan expired, or is the loan guaranteed. Including whether it has been done. Shown are total score, lien priority, outstanding payable balance (unit: 1 million), total unpaid principal%, and cash flow analysis.

The method and system for the valuation of assets using the fuzzy clustering system 50 includes a list of executable instructions for performing logical functions. This list is by or related to instruction execution systems, equipment or devices, such as computer-based systems, embedded processors, or other systems capable of fetching and executing instructions from an instruction execution system, equipment, or device. Can be realized in a computer-readable medium used in the following. In the context of this document, "computer-readable medium" means any means capable of embedding, storing, transmitting, propagating or transferring a program used by or in connection with an instruction execution system, device or apparatus.

Computer readable media may be, for example (but not limited to) electrical, magnetic,
It may be an optical, electromagnetic, infrared or semiconductor system, equipment, device or means of propagation.
More specific examples (non-exhaustive list) of computer readable media include: Electrical connection (electronic) with one or more wires, movable computer disk (magnetic), random access memory (RAM) (magnetic), read only memory (ROM) (magnetic), erasable programmable read only memory (EPROM
Or flash memory) (magnetic), optical fiber (optical), and moving compact disk read only memory (CDROM) (optical).

It should be noted that the computer-readable medium may be a paper on which the program is printed or another appropriate medium. This means that the program is electronically captured, for example by optical scanning of paper or other media, compiled, recognized or, if necessary, processed in a suitable manner and then stored in computer memory. This is because it is possible to

The above description has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Obvious modifications or variations are possible in light of the above teachings. The flowchart of the present disclosure illustrates the architecture, functionality and operation of a possible implementation of a registry usage optimization, editing, and translation system. In this regard, each block represents a module, segment, or portion of code that contains one or more executable instructions for implementing the specified logical function. Also, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures,
Or, for example, in practice, it may occur at about the same time or in reverse order, depending on the functions involved.

The above-described systems and methods will enable those of ordinary skill in the art to use the invention in various embodiments suitable for the particular use envisioned, and with various modifications, as well as optimal illustrations of the principles of the invention. It has been selected and described in order to provide the best example with practical application. All such modifications and variations are intended to be within the scope of the invention as determined by the following claims when interpreted according to their fair and lawful scope.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating an example of valuing assets using a fuzzy clustering system, using an example loan portfolio.

FIG. 2 is a block diagram illustrating the valuation of assets in a computer readable medium in a computer system using fuzzy clustering.

FIG. 3 is a block diagram illustrating an example process flow of a system and method for asset valuation using fuzzy clustering.

FIG. 4 is a flow chart illustrating an example process flow of a system and method for valuing assets using fuzzy clustering for different types of loan portfolio examples.

5 is a block diagram of an example showing different types of data acquired during a data acquisition process, such as those shown in FIGS. 2, 3 and 4. FIG.

FIG. 6 is a table of examples of critical variables and coding schemes used in the hierarchical segmentation process as shown in FIGS. 3 and 4.

FIG. 7 is a block diagram of an example showing the use of segmentation modeling for an example loan portfolio as shown in FIGS. 2, 3 and 4 above.

FIG. 8 is an example flow chart of a fuzzy clustering process used in the system and method for valuing assets of the present invention, for example loan portfolios as shown in FIGS. 2, 3 and 4. Is.

9 is a process for calculating a fuzzy clustering average in the asset valuation system and method of the present invention for loan and portfolio examples as shown in FIGS. 2, 3, 4 and 8. FIG. 3 is a flowchart of an example of FIG.

FIG. 10 is a flow chart of an example underwriting review process in the system and method of valuing assets of the present invention in the example of a loan in a portfolio, as shown in FIGS. 2, 3 and 4.

FIG. 11A is a diagram showing an example of intra-cluster dispersion of 6 clusters created in a dendrogram as shown in FIG. 11D.

FIG. 11B is a diagram showing intercluster dispersion of 6 clusters, as shown in FIGS. 11A and 11C.

FIG. 11C is a diagram called a dendrogram showing each pair of cluster centroids using a distance matrix as shown in FIG.

FIG. 11D is a diagram of a tree diagram created during reproduction of the tree creation diagram as shown in FIGS. 8 and 11B.

FIG. 12A   FIG. 12 is an example HELTR table for evaluation of 6 clusters, as shown in FIG. 11D.

FIG. 12B   FIG. 12 is an example HELTR table for evaluation of 6 clusters, as shown in FIG. 11D.

─────────────────────────────────────────────────── ─── Continued front page    (81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, I T, LU, MC, NL, PT, SE, TR), OA (BF , BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, G M, KE, LS, MW, MZ, SD, SL, SZ, TZ , UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, C N, CU, CZ, DE, DK, EE, ES, FI, GB , GD, GE, GH, GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP, KR, KZ, L C, LK, LR, LS, LT, LU, LV, MD, MG , MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, T J, TM, TR, TT, UA, UG, US, UZ, VN , YU, ZW (72) Inventor Johnson, Christopher Donald             United States, 12065, New York,             Clifton Park, Berkshire Dora             Eve West, number 17 (72) Inventor Keys, Tim Kelly             United States, 06896, Connecticut,             West Reading, Top Ledge             No.16 (72) Inventor Pispathi, Chandrasekhar             United States, 12309, New York,             Niskayuna, apartment number 5             Bee 49, Hillside Avenue, 1187 (72) Inventor Stuart, William Cree             United States, 06855, Connecticut,             Norwalk, Sycamore Street, 13             Turn

Claims (21)

[Claims] 1. A method for providing asset valuation using fuzzy clustering, wherein a plurality of assets are acquired and a plurality of variables are attached to the plurality of assets. Inspecting the plurality of variables using a model, calculating values of the plurality of assets and the variables using fuzzy clustering, and calculating profits of the plurality of assets, How to include. 2. The step of using fuzzy clustering randomizes a membership degree of each of a plurality of data points of a plurality of clusters to obtain the plurality of data points and the plurality of data points. The method of claim 1, further comprising calculating a centroid with the cluster. 3. The method of claim 2, wherein the step of using fuzzy clustering further comprises updating a respective degree of membership of the plurality of data points of each of the plurality of clusters. 4. The method of claim 3, wherein the step of using fuzzy clustering further comprises: calculating a value using an objective function and determining if the objective function value converges. 5. The step of using fuzzy clustering comprises:   Calculate the intercluster variance,   The method of claim 2, further comprising calculating an intra-cluster variance. 6. The method of claim 5, wherein the step of using fuzzy clustering further comprises compacting the intercluster variance and the intracluster variance. 7. The step of compacting the inter-cluster dispersion and the intra-cluster dispersion calculates a distance between each pair of the inter-cluster dispersion and the intra-cluster dispersion, and the inter-cluster dispersion and the inter-cluster dispersion are calculated. Store the distance between each pair with the intra-cluster variance in a distance matrix, create a dendrogram for the distance between each pair of the inter-cluster variance and the intra-cluster variance, and 7. The method of claim 6, further comprising evaluating the dendrogram for finding. 8. A system for providing asset valuation using fuzzy clustering, comprising means for acquiring a plurality of assets, means for attaching a plurality of variables to the plurality of assets, and a plurality of the plurality of assets. Means for inspecting the assets and the plurality of variables using a model, means for calculating the values of the plurality of assets and the variables using fuzzy clustering, and the plurality of assets And means for calculating the profit of the. 9. The means for using fuzzy clustering comprises means for randomizing each membership degree of each of the plurality of data points of the plurality of clusters, the plurality of data points and the plurality of data points. 9. The system of claim 8, further comprising means for calculating a centroid with a plurality of clusters. 10. The system of claim 9, wherein the means for using fuzzy clustering further comprises means for updating respective membership degrees of the plurality of data points of each of the plurality of clusters. 11. The system of claim 10, wherein the means for using fuzzy clustering further comprises: means for calculating a value using an objective function; and means for determining if the value converges. 12. The means for using fuzzy clustering comprises:   Means for calculating the inter-cluster variance,   9. The system of claim 8, further comprising means for calculating intra-cluster variance. 13. The system of claim 12, wherein the means for using fuzzy clustering further comprises means for compacting the intercluster dispersion and the intracluster dispersion. 14. The means for using fuzzy clustering means comprises means for calculating a distance between each pair of the inter-cluster dispersion and the intra-cluster dispersion; and the inter-cluster dispersion and the intra-cluster dispersion. Means for storing the distances between each pair in a distance matrix, means for creating a dendrogram for the distances between each pair of the inter-cluster variance and the intra-cluster variance, and finding possible integrations 14. The system of claim 13, further comprising means for evaluating the dendrogram for. 15. A system for providing stability analysis of profits of different types of goods or service products, wherein the data acquisition logic acquires a plurality of assets and attaches a plurality of variables to the plurality of assets. Segmentation logic for inspecting the plurality of assets and the plurality of variables using a model; fuzzy clustering logic for calculating values of the plurality of assets and the variables; And a profit analysis logic that calculates the profit of each asset. 16. The fuzzy clustering logic includes: logic for randomizing a membership degree of each of a plurality of data points of a plurality of clusters; and the plurality of data points and the plurality of data points. 16. The system of claim 15, further comprising logic to calculate a centroid with the cluster of. 17. The system of claim 16, further comprising logic for updating a respective degree of membership of each of the plurality of data points of each of the plurality of clusters. 18. The system of claim 17, further comprising logic to calculate a value using an objective function and determine if the value converges. 19. Logic for calculating an inter-cluster variance from the plurality of data of each of the plurality of clusters and an intra-cluster variance from the plurality of data points of each of the plurality of clusters. The system of claim 17, further comprising logic. 20. The system of claim 19, further comprising logic for compacting the inter-cluster variance and the intra-cluster variance. 21. Logic for calculating a distance between each pair of the inter-cluster dispersion and the intra-cluster dispersion, and storing a distance between each pair of the inter-cluster dispersion and the intra-cluster dispersion in a distance matrix. Logic to create a dendrogram for the distance between each pair of the inter-cluster variance and the intra-cluster variance, and logic to evaluate the dendrogram to find possible integrations. The system of claim 16, further comprising: