Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
6Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1 Assignment
0 Petitions
Accused Products
Abstract
A data analyzer/classifier comprises using a preprocessing step, and energy minimization step, and a postprocessing step to analyze/classify data.
44 Citations
View as Search Results
Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 7,702,155 B2
Filed 04/01/2005

Current Assignee
Abel Wolman, Glickman Jeff B.

Sponsoring Entity
Abel Wolman, Glickman Jeff B.

Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 7,272,262 B2
Filed 04/01/2005

Current Assignee
Abel Wolman, Glickman Jeff B.

Sponsoring Entity
Abel Wolman, Glickman Jeff B.

Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 20050175244A1
Filed 04/01/2005

Current Assignee
Abel Wolman, Glickman Jeff B.

Sponsoring Entity
Abel Wolman, Glickman Jeff B.

Insight and algorithmic clustering for automated synthesis  
Patent #
US 9,336,302 B1
Filed 03/14/2013

Current Assignee
OOL LLC

Sponsoring Entity
OOL LLC

Insight and algorithmic clustering for automated synthesis  
Patent #
US 9,607,023 B1
Filed 05/06/2016

Current Assignee
OOL LLC

Sponsoring Entity
OOL LLC

Insight and algorithmic clustering for automated synthesis  
Patent #
US 10,318,503 B1
Filed 03/27/2017

Current Assignee
OOL LLC

Sponsoring Entity
OOL LLC

Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 7,174,048 B2
Filed 03/31/2005

Current Assignee
Abel Wolman, Glickman Jeff B.

Sponsoring Entity
Abel Wolman, Glickman Jeff B.

Method, system and computer program product for nonlinear mapping of multidimensional data  
Patent #
US 6,571,227 B1
Filed 05/03/1999

Current Assignee
3Dimensional Pharmaceuticals Inc.

Sponsoring Entity
3Dimensional Pharmaceuticals Inc.

Method of generating a browser interface for representing similarities between segments of code  
Patent #
US 5,574,837 A
Filed 01/17/1995

Current Assignee
AlcatelLucent USA Inc.

Sponsoring Entity
AlcatelLucent USA Inc.

System, method, and computer program product for representing object relationships in a multidimensional space  
Patent #
US 7,039,621 B2
Filed 03/22/2001

Current Assignee
Johnson Johnson Pharmaceutical Research Development LLC

Sponsoring Entity
Johnson Johnson Pharmaceutical Research Development LLC

Method for generating analyses of categorical data  
Patent #
US 6,714,897 B2
Filed 01/02/2001

Current Assignee
Battelle Memorial Institute

Sponsoring Entity
Battelle Memorial Institute

Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 6,993,186 B1
Filed 12/23/1998

Current Assignee
GLICKMAN JEFF B., Abel Wolman

Sponsoring Entity
GLICKMAN JEFF B., Abel Wolman

Apparatus for producing a flattening map of a digitized image for conformally mapping onto a surface and associated method  
Patent #
US 6,697,538 B1
Filed 07/28/2000

Current Assignee
Brigham and Womens Hospital Incorporated

Sponsoring Entity
Brigham and Womens Hospital Incorporated

Method, system, and computer program product for analyzing combinatorial libraries  
Patent #
US 7,054,757 B2
Filed 01/29/2002

Current Assignee
Johnson Johnson Pharmaceutical Research Development LLC

Sponsoring Entity
Johnson Johnson Pharmaceutical Research Development LLC

Method and system for interactive parameter optimization using multidimensional scaling  
Patent #
US 20060173874A1
Filed 01/27/2006

Current Assignee
Siemens Medical Solutions USA Incorporated

Sponsoring Entity
Siemens Medical Solutions USA Incorporated

Node localization in communication networks  
Patent #
US 20050080924A1
Filed 10/09/2003

Current Assignee
Palo Alto Research Center Inc.

Sponsoring Entity
Palo Alto Research Center Inc.

Energy minimization for classification, pattern recognition, sensor fusion, data compression, network reconstruction and signal processing  
Patent #
US 20050175244A1
Filed 04/01/2005

Current Assignee
Abel Wolman, Glickman Jeff B.

Sponsoring Entity
Abel Wolman, Glickman Jeff B.

Threedimensional face recognition  
Patent #
US 6,947,579 B2
Filed 10/31/2002

Current Assignee
Intel Benelux BV

Sponsoring Entity
TECHNION RESEARCH AND DEVELOPMENT FOUNDATION LTD.

Energy minimization for data merging and fusion  
Patent #
US 6,968,342 B2
Filed 06/19/2001

Current Assignee
GLICKMAN JEFF B., Abel Wolman

Sponsoring Entity
GLICKMAN JEFF B., Abel Wolman

Detection and classification system for analyzing deterministic properties of data using correlation parameters  
Patent #
US 6,732,064 B1
Filed 05/13/2002

Current Assignee
Adobe Systems Incorporated

Sponsoring Entity
Nonlinear Solutions Inc.

Information retrieving system and method  
Patent #
US 6,804,420 B2
Filed 09/21/2001

Current Assignee
Fujitsu Limited

Sponsoring Entity
Fujitsu Limited

Geometrization for pattern recognition, data analysis, data merging, and multiple criteria decision making  
Patent #
US 20040230586A1
Filed 07/30/2003

Current Assignee
Abel Wolman

Sponsoring Entity


Video object segmentation using active contour modelling with global relaxation  
Patent #
US 6,546,117 B1
Filed 06/10/1999

Current Assignee
University of Washington

Sponsoring Entity
University of Washington

Energy minimization for data merging and fusion  
Patent #
US 20020002555A1
Filed 06/19/2001

Current Assignee
GLICKMAN JEFF B., Abel Wolman

Sponsoring Entity
GLICKMAN JEFF B., Abel Wolman

Image processing method including steps for the segmentation of a multidimensional image, and medical imaging apparatus utilizing this method  
Patent #
US 6,332,034 B1
Filed 03/19/1999

Current Assignee
US Philips Corporation

Sponsoring Entity
US Philips Corporation

Computerassisted method and apparatus for the detection of lung nodules  
Patent #
US 5,987,094 A
Filed 10/30/1997

Current Assignee
University of South Florida

Sponsoring Entity
University of South Florida

Method of modifying a timevarying image sequence by estimation of velocity vector fields  
Patent #
US 5,991,459 A
Filed 01/22/1992

Current Assignee
Eastman Kodak Company

Sponsoring Entity
Eastman Kodak Company

Blind signal processing system employing information maximization to recover unknown signals through unsupervised minimization of output redundancy  
Patent #
US 5,706,402 A
Filed 11/29/1994

Current Assignee
Salk Institute for Biological Studies

Sponsoring Entity
Salk Institute for Biological Studies

Apparatus and method for fusing diverse data  
Patent #
US 5,729,451 A
Filed 12/01/1995

Current Assignee
CyTerra Corp.

Sponsoring Entity
Coleman Research Corp.

System and process for constructing optimized prototypes for pattern recognition using competitive classification learning  
Patent #
US 5,802,207 A
Filed 06/30/1995

Current Assignee
Industrial Technology Research Institute

Sponsoring Entity
Industrial Technology Research Institute

Method and apparatus for efficient presentation of highquality threedimensional audio  
Patent #
US 5,596,644 A
Filed 10/27/1994

Current Assignee
Creative Technology Ltd.

Sponsoring Entity
AUREAL SEMICONDUCTOR INC.

Method of generating dictionary for pattern recognition and pattern recognition method using the same  
Patent #
US 5,602,938 A
Filed 05/20/1994

Current Assignee
Nippon Telegraph and Telephone Corporation

Sponsoring Entity
Nippon Telegraph and Telephone Corporation

Method and system for twodimensional visualization of an information taxonomy and of text documents based on topical content of the documents  
Patent #
US 5,625,767 A
Filed 03/13/1995

Current Assignee
Robert Clarke, Bartell Brian

Sponsoring Entity
Robert Clarke, Bartell Brian

Method and system to enhance medical signals for realtime analysis and highresolution display  
Patent #
US 5,490,516 A
Filed 05/06/1993

Current Assignee
William H. Hutson

Sponsoring Entity
William H. Hutson

Method of predicting carcinomic metastases  
Patent #
US 5,579,766 A
Filed 07/31/1995

Current Assignee
Board of Regents of the University of Texas System

Sponsoring Entity
Board of Regents of the University of Texas System

Twostep method constructing largearea facilities and smallarea intrafacilities equipments optimized by user population density  
Patent #
US 5,402,335 A
Filed 09/24/1992

Current Assignee
United States Of America As Represented By The Secretary Of The Navy

Sponsoring Entity
United States Of America As Represented By The Secretary Of The Navy

Apparatus and method for improving recognition of patterns by prototype transformation  
Patent #
US 5,422,961 A
Filed 04/03/1992

Current Assignee
ATT Inc.

Sponsoring Entity
ATT Inc.

Method of predicting carcinomic metastases  
Patent #
US 5,437,279 A
Filed 07/02/1992

Current Assignee
Board of Regents of the University of Texas System

Sponsoring Entity
Board of Regents of the University of Texas System

Data fusion workstation  
Patent #
US 5,321,613 A
Filed 11/12/1992

Current Assignee
CyTerra Corp.

Sponsoring Entity
Coleman Research Corp.

Method and system for near realtime analysis and display of electrocardiographic signals  
Patent #
US 5,348,020 A
Filed 12/15/1992

Current Assignee
TSA AMERICA INC. A CORP. OF DELAWARE

Sponsoring Entity
William H. Hutson

General method of pattern classification using the two domain theory  
Patent #
US 5,181,259 A
Filed 09/25/1990

Current Assignee
The United States of America As Represented By The Secretary of Agriculture

Sponsoring Entity
The United States of America As Represented By The Secretary of Agriculture

Process which aids to the laying out of locations of a limited number of 100, personnel and equipments in functional organization  
Patent #
US 5,235,506 A
Filed 08/30/1991

Current Assignee
United States Of America As Represented By The Secretary Of The Navy

Sponsoring Entity
United States Of America As Represented By The Secretary Of The Navy

Multidimensional signal processing and display  
Patent #
US 5,245,587 A
Filed 11/18/1992

Current Assignee
William H. Hutson

Sponsoring Entity
William H. Hutson

Multidimensional data processing and display  
Patent #
US 5,175,710 A
Filed 12/14/1990

Current Assignee
William H. Hutson

Sponsoring Entity
William H. Hutson

0 Claims
1 Specification
The present patent document is a division of application Ser. No. 09/581,949, filed Jun. 19, 2000, which is a continuation of International Application number PCT/US98/27374, filed Dec. 23, 1998, which claims the benefit of the filing date under 35 U.S.C. § 119(e) of Provisional U.S. patent application Ser. 60/071,592, filed Dec. 29, 1997. All of the foregoing applications are hereby incorporated by reference.
A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
A computer program listing appendix is included containing computer program code listings on a CDRom pursuant to 37 C.F.R. 1.52(e) and is hereby incorporated herein by reference in its entirety. The total number of compact discs is 1 (two duplicate copies are filed herewith). Each compact disc includes one (1) file entitled SOURCE CODE APPENDIX. Each compact disc includes 23,167 bytes. The creation date of the compact disc is Mar. 24, 2005.
The present invention relates to recognition, analysis, and classification of patterns in data from real world sources, events and processes. Patterns exist throughout the real world. Patterns also exist in the data used to represent or convey or store information about real world objects or events or processes. As information systems process more real world data, there are mounting requirements to build more sophisticated, capable and reliable pattern recognition systems.
Existing pattern recognition systems include statistical, syntactic and neural systems. Each of these systems has certain strengths which lends it to specific applications. Each of these systems has problems which limit its effectiveness.
Some real world patterns are purely statistical in nature. Statistical and probabilistic pattern recognition works by expecting data to exhibit statistical patterns. Pattern recognition by this method alone is limited. Statistical pattern recognizers cannot see beyond the expected statistical pattern. Only the expected statistical pattern can be detected.
Syntactic pattern recognizers function by expecting data to exhibit structure. While syntactic pattern recognizers are an improvement over statistical pattern recognizers, perception is still narrow and the system cannot perceive beyond the expected structures. While some real world patterns are structural in nature, the extraction of structure is unreliable.
Pattern recognition systems that rely upon neural pattern recognizers are an improvement over statistical and syntactic recognizers. Neural recognizers operate by storing training patterns as synaptic weights. Later stimulation retrieves these patterns and classifies the data. However, the fixed structure of neural pattern recognizers limits their scope of recognition. While a neural system can learn on its own, it can only find the patterns that its fixed structure allows it to see. The difficulties with this fixed structure are illustrated by the wellknown problem that the number of hidden layers in a neural network strongly affects its ability to learn and generalize. Additionally, neural pattern recognition results are often not reproducible. Neural nets are also sensitive to training order, of ten require redundant data for training, can be slow learners and sometimes never learn. Most importantly, as with statistical and syntactic pattern recognition systems, neural pattern recognition systems are incapable of discovering truly new knowledge.
Accordingly, there is a need for an improved method and apparatus for pattern recognition, analysis, and classification which is not encumbered by preconceptions about data or models.
By way of illustration only, an analyzer/classifier process for data comprises using energy minimization with one or more input matrices. The data to be analyzed/classified is processed by an energy minimization technique such as individual differences multidimensional scaling (IDMDS) to produce at least a rate of change of stress/energy. Using the rate of change of stress/energy and possibly other IDMDS output, the data are analyzed or classified through patterns recognized within the data. The foregoing discussion of one embodiment has been presented only by way of introduction. Nothing in this section should be taken as a limitation on the following claims, which define the scope of the invention.
The method and apparatus in accordance with the present invention provide an analysis tool with many applications. This tool can be used for data classification, pattern recognition, signal processing, sensor fusion, data compression, network reconstruction, and many other purposes. The invention relates to a general method for data analysis based on energy minimization and least energy deformations. The invention uses energy minimization principles to analyze one to many data sets. As used herein, energy is a convenient descriptor for concepts which are handled similarly mathematically. Generally, the physical concept of energy is not intended by use of this term but the more general mathematical concept. Within multiple data sets, individual data sets are characterized by their deformation under least energy merging. This is a contextual characterization which allows the invention to exhibit integrated unsupervised learning and generalization. A number of methods for producing energy minimization and least energy merging and extraction of deformation information have been identified; these include, the finite element method (FEM), simulated annealing, and individual differences multidimensional scaling (IDMDS). The presently preferred embodiment of the invention utilizes individual differences multidimensional scaling (IDMDS).
Multidimensional scaling (MDS) is a class of automated, numerical techniques for converting proximity data into geometric data. IDMDS is a generalization of MDS, which converts multiple sources of proximity data into a common geometric configuration space, called the common space, and an associated vector space called the source space. Elements of the source space encode deformations of the common space specific to each source of proximity data. MDS and IDMDS were developed for psychometric research, but are now standard tools in many statistical software packages. MDS and IDMDS are often described as data visualization techniques. This description emphasizes only one aspect of these algorithms.
Broadly, the goal of MDS and IDMDS is to represent proximity data in a low dimensional metric space. This has been expressed mathematically by others (see, for example, de Leeuw, J. and Heiser, W., “Theory of multidimensional scaling,” in P. R. Krishnaiah and L. N. Kanal, eds., Handbook of Statistics, Vol. 2. NorthHolland, New York, 1982) as follows. Let S be a nonempty finite set, p a real valued function on S×S,
p:S×S→R.
p is a measure of proximity between objects in S. Then the goal of MDS is to construct a mapping {tilde over (f)} from S into a metric space (X, d),
{tilde over (f)}:S→X,
such that p(i, j)=p_{ij}≈d({tilde over (f)}(i), {tilde over (f)}(j)), that is, such that the proximity of object i to object j in S is approximated by the distance in X between {tilde over (f)}(i) and {tilde over (f)}(j). X is usually assumed to be n dimensional Euclidean space R^{n}, with n sufficiently small.
IDMDS generalizes MDS by allowing multiple sources. For k=1, . . . , m let S_{k }be a finite set with proximity measure p_{k}, then IDMDS constructs maps
{tilde over (f)}_{k}:S_{k}→X
such that p_{k}(i, j)=p_{ijk}≈d({tilde over (f)}_{k}(i), {tilde over (f)}_{k}(j)), for k=1, . . . , m.
Intuitively, IDMDS is a method for representing many points of view. The different proximities p_{k }can be viewed as giving the proximity perceptions of different judges. IDMDS accommodates these different points of view by finding different maps {tilde over (f)}_{k }for each judge. These individual maps, or their image configurations, are deformations of a common configuration space whose interpoint distances represent the common or merged point of view.
MDS and IDMDS can equivalently be described in terms of transformation functions. Let P=(p_{ij}) be the matrix defined by the proximity p on S×S. Then MDS defines a transformation function
f:p_{ij}d_{ij}(X),
where d_{ij}(X)=d({tilde over (f)}(i), {tilde over (f)}(j)), with {tilde over (f)} the mapping from S→X induced by the transformation function f. Here, by abuse of notation, X={tilde over (f)}(S), also denotes the image of S under {tilde over (f)}. The transformation function f should be optimal in the sense that the distances f(p_{ij}) give the best approximation to the proximities p_{ij}. This optimization criterion is described in more detail below. IDMDS is similarly reexpressed; the single transformation f is replaced by m transformations f_{k}. Note, these f_{k }need not be distinct. In the following, the image of S_{k }under f_{k }will be written X_{k}.
MDS and IDMDS can be further broken down into socalled metric and nonmetric versions. In metric MDS or IDMDS, the transformations f(f_{k}) are parametric functions of the proximities p_{ij}(p_{ijk}). Nonmetric MDS or IDMDS generalizes the metric approach by allowing arbitrary admissible transformations f (f_{k}), where admissible means the association between proximities and transformed proximities (also called disparities in this context) is weakly monotone:
p_{ij}<p_{kl }implies f(p_{ij})≦f(p_{kl}).
Beyond the metricnonmetric distinction, algorithms for MDS and IDMDS are distinguished by their optimization criteria and numerical optimization routines. One particularly elegant and publicly available IDMDS algorithm is PROXSCAL See Commandeur, J. and Heiser, W., “Mathematical derivations in the proximity scaling (PROXSCAL) of symmetric data matrices,” Tech. report no. RR9303, Department of Data Theory, Leiden University, Leiden, The Netherlands. PROXSCAL is a least squares, constrained majorization algorithm for IDMDS. We now summarize this algorithm, following closely the above reference.
PROXSCAL is a least squares approach to IDMDS which minimizes the objective function
σ is called the stress and measures the goodnessoffit of the configuration distances d_{ij}(X_{k}) to the transformed proximities f_{k}(p_{ijk}). This is the most general form for the objective function. MDS can be interpreted as an energy minimization process and stress can be interpreted as an energy functional. The w_{ijk }are proximity weights. For simplicity, it is assumed in what follows that w_{ijk}=1 for all i, j, k.
The PROXSCAL majorization algorithm for MDS with transformations is summarized as follows.
1. Choose a (possibly random) initial configuration X^{0}.
2. Find optimal transformations f(p_{ij}) for fixed distances d_{ij}(X^{0}).
3. Compute the initial stress
4. Compute the Guttman transform {overscore (X)} of X^{0 }with the transformed proximities f(p_{ij}). This is the majorization step.
5. Replace X^{0 }with X and find optimal transformations f(p_{ij}) for fixed distances d_{ij}(X^{0}).
6. Compute σ(f,X^{0}).
7.Go to step 4 if the difference between the current and previous stress is not less than ε, some previously defined number. Stop otherwise.
For multiple sources of proximity data, restrictions are imposed on the configurations X_{k }associated to each source of proximity data in the form of the constraint equation X_{k}=ZW_{k}.
This equation defines a common configuration space Z and diagonal weight matrices W_{k}. Z represents a merged or common version of the input sources, while the W_{k }define the deformation of the common space required to produce the individual configurations X_{k}. The vectors defined by diag(W_{k}), the diagonal entries of the weight matrices W_{k}, form the source space W associated to the common space Z.
The PROXSCAL constrained majorization algorithm for IDMDS with transformations is summarized as follows. To simplify the discussion, socalled unconditional IDMDS is described. This means the m transformation functions are the same: f_{1}=f_{2}=. . . =f_{m}.
1. Choose constrained initial configurations X_{k}^{0}.
2. Find optimal transformations f(p_{ijk}) for fixed distances d_{ij}(X_{k}^{0}).
3. Compute the initial stress
4. Compute unconstrained updates {overscore (X)}_{k }of X_{k}^{0 }using the Guttman transform with transformed proximities f(p_{ijk}) This is the unconstrained majorization step.
5. Solve the metric projection problem by finding X_{k}^{+} minimizing
subject to the constraints X_{k}=ZW_{k}. This step constrains the updated configurations from step 4.
6. Replace X_{k}^{0 }with X_{k}^{+} and find optimal transformations f(p_{ijk}) for fixed distances d_{ij}(X_{k}^{0}).
7. Compute σ(f, X_{1}^{0}, . . . , X_{m}^{0}).
8. Go to step 4 if the difference between the current and previous stress is not less than ε, some previously defined number. Stop otherwise.
Here, tr(A) and A′ denote, respectively, the trace and transpose of matrix A.
43 It should be pointed out that other IDMDS routines do not contain an explicit constraint condition. For example, ALSCAL (see Takane, Y., Young, F, and de Leeuw, J., “Nonmetric individual differences multidimensional scaling: an alternating least squares method with optimal scaling features,” Psychometrika, Vol. 42, 1977) minimizes a different energy expression (sstress) over transformations, configurations, and weighted Euclidean metrics. ALSCAL also produces common and source spaces, but these spaces are computed through alternating least squares without explicit use of constraints. Either form of IDMDS can be used in the present invention.
MDS and IDMDS have proven useful for many kinds of analyses. However, it is believed that prior utilizations of these techniques have not extended the use of these techniques to further possible uses for which MDS and IDMDS have particular utility and provide exceptional results. Accordingly, one benefit of the present invention is to incorporate MDS or IDMDS as part of a platform in which aspects of these techniques are extended. A further benefit is to provide an analysis technique, part of which uses IDMDS, that has utility as an analytic engine applicable to problems in classification, pattern recognition, signal processing, sensor fusion, and data compression, as well as many other kinds of data analytic applications.
Referring now to
It is to be understood that the steps forming the tool 100 may be implemented in a computer usable medium or in a computer system as computer executable software code. In such an embodiment, step 110 may be configured as a code, step 120 may be configured as second code and step 120 may be configured as third code, with each code comprising a plurality of machine readable steps or operations for performing the specified operations. While step 110, step 120 and step 130 have been shown as three separate elements, their functionality can be combined and/or distributed. It is to be further understood that “medium” is intended to broadly include any suitable medium, including analog or digital, hardware or software, now in use or developed in the future.
Step 110 of the tool 100 is the transformation of the data into matrix form. The only constraint on this transformation for the illustrated embodiment is that the resulting matrices be square. The type of transformation used depends on the data to be processed and the goal of the analysis. In particular, it is not required that the matrices be proximity matrices in the traditional sense associated with IDMDS. For example, time series and other sequential data may be transformed into source matrices through straight substitution into entries of symmetric matrices of sufficient dimensionality (this transformation will be discussed in more detail in an example below). Time series or other signal processing data may also be Fourier or otherwise analyzed and then transformed to matrix form.
Step 120 of the tool 100 implements energy minimization and extraction of deformation information through IDMDS. In the IDMDS embodiment of the tool 100, the stress function σ defines an energy functional over configurations and transformations. The configurations are further restricted to those which satisfy the constraint equations X_{k}=ZW_{k}. For each configuration X_{k}, the weight vectors diag(W_{k}) are the contextual signature, with respect to the common space, of the kth input source. Interpretation of σ as an energy functional is fundamental; it greatly expands the applicability of MDS as an energy minimization engine for data classification and analysis.
Step 130 consists of both visual and analytic methods for decoding and interpreting the source space W from step 120. Unlike traditional applications of IDMDS, tool 100 often produces high dimensional output. Among other things, this makes visual interpretation and decoding of the source space problematic. Possible analytic methods for understanding the high dimensional spaces include, but are not limited to, linear programming techniques for hyperplane and decision surface estimation, cluster analysis techniques, and generalized gravitational model computations. A source space dyedropping or tracer technique has been developed for both source space visualization and analytic postprocessing. Step 130 may also consist in recording stress/energy, or the rate of change of stress/energy, over multiple dimensions. The graph of energy (rate or change or stress/energy) against dimension can be used to determine network and dynamical system dimensionality. The graph of stress/energy against dimensionality is traditionally called a scree plot. The use and purpose of the scree plot is greatly extended in the present embodiment of the tool 100.
Let S={S_{k}} be a collection of data sets or sources S_{k }for k=1, . . . , m Step 110 of the tool 100 converts each S_{k}∈S to matrix form M(S_{k}) where M(S_{k}) is a p dimensional real hollow symmetric matrix. Hollow means the diagonal entries of M(S_{k}) are zero. As indicated above, M(S_{k}) need not be symmetric or hollow, but for simplicity of exposition these additional restrictions are adopted. Note also that the matrix dimensionality p is a function of the data S and the goal of the analysis. Since M(S_{k}) is hollow symmetric, it can be interpreted and processed in IDMDS as a proximity (dissimilarity) matrix. Step 110 can be represented by the map
M:S→H^{p}(R),
S_{kM(S}_{k})
where H^{p}(R) is the set of p dimensional hollow real symmetric matrices. The precise rule for computing M depends on the type of data in S, and the purpose of the analysis. For example, if S contains time series data, then M might entail the straightforward entrywise encoding mentioned above. If S consists of optical character recognition data, or some other kind of geometric data, then M(S_{k}) may be a standard distance matrix whose ijth entry is the Euclidean distance between “on” pixels i and j. M can also be combined with other transformations to form the composite, (M∘F)(S_{k}), where F, for example, is a fast Fourier transform (FFT) on signal data S_{k}. To make this more concrete, in the examples below M will be explicitly calculated in a number of different ways. It should also be pointed out that for certain data collections S it is possible to analyze the conjugate or transpose S^{l }of S. For instance, in data mining applications, it is useful to transpose records (clients) and fields (client attributes) thus allowing analysis of attributes as well as clients. The mapping M is simply applied to the transposed data.
Step 120 of the presently preferred embodiment of the tool 100 is the application of IDMDS to the set of input matrices M(S)={M(S_{k})}. Each M(S_{k})∈M(S) is an input source for IDMDS. As described above, the IDMDS output is a common space Z⊂R^{n }and a source space W. The dimensionality n of these spaces depends on the input data S and the goal of the analysis. For signal data, it is often useful to set n=p−1 or even n=S_{k} where S_{k} denotes the cardinality of S_{k}. For data compression, low dimensional output spaces are essential. In the case of network reconstruction, system dimensionality is discovered by the invention itself.
IDMDS can be thought of as a constrained energy minimization process. As discussed above, the stress σ is an energy functional defined over transformations and configurations in R^{n}; the constraints are defined by the constraint equation X_{k}=ZW_{k}. IDMDS attempts to find the lowest stress or energy configurations X_{k }which also satisfy the constraint equation. (MDS is the special case when each W_{k}=I, the identity matrix.) Configurations X_{k }most similar to the source matrices M(S_{k}) have the lowest energy. At the same time, each X_{k }is required to match the common space Z up to deformation defined by the weight matrices W_{k}. The common space serves as a characteristic, or reference object. Differences between individual configurations are expressed in terms of this characteristic object with these differences encoded in the weight matrices W_{k}. The deformation information contained in the weight matrices, or, equivalently, in the weight vectors defined by their diagonal entries, becomes the signature of the configurations X_{k }and hence the sources S_{k }(through M(S_{k})). The source space may be thought of as a signature classification space.
The weight space signatures are contextual; they are defined with respect to the reference object Z. The contextual nature of the source deformation signature is fundamental. As the polygon classification example below will show, Zcontextuality of the signature allows the tool 100 to display integrated unsupervised machine learning and generalization. The analyzer/classifier learns seamlessly and invisibly. Zcontextuality also allows the tool 100 to operate without a priori data models. The analyzer/classifier constructs its own model of the data, the common space Z.
Step 130, the back end of the tool 100, decodes and interprets the source or classification space output W from IDMDS. Since this output can be high dimensional, visualization techniques must be supplemented by analytic methods of interpretation. A dyedropping or tracer technique has been developed for both visual and analytic postprocessing. This entails differential marking or coloring of source space output. The specification of the dyedropping is contingent upon the data and overall analysis goals. For example, dyedropping may be twocolor or binary allowing separating hyperplanes to be visually or analytically determined. For an analytic approach to separating hyperplanes using binary dyedropping see Bosch, R. and Smith, J, “Separating hyperplanes and the authorship of the disputed federalist papers,” American Mathematical Monthly, Vol. 105, 1998. Discrete dyedropping allows the definition of generalized gravitational clustering measures of the form
Here, A denotes a subset of W (indicated by dyedropping), X_{A}(x), is the characteristic function on A, d (. , .) is a distance function, and p∈R. Such measures may be useful for estimating missing values in data bases. Dyedropping can be defined continuously, as well, producing a kind of height function on W. This allows the definition of decision surfaces or volumetric discriminators. The source space W is also analyzable using standard cluster analytic techniques. The precise clustering metric depends on the specifications and conditions of the IDMDS analysis in question.
Finally, as mentioned earlier, the stress/energy and rate of change of stress/energy can be used as postprocessing tools. Minima or kinks in a plot of energy, or the rate of change of energy, over dimension can be used to determine the dimensionality of complex networks and general dynamical systems for which only partial output information is available. In fact, this technique allows dimensionality to be inferred often from only a single data stream of time series of observed data.
A number of examples are presented below to illustrate the method and apparatus in accordance with the present invention. These examples are illustrative only and in no way limit the scope of the method or apparatus.
The goal of this experiment was to classify a set of regular polygons. The collection S={S_{1}, . . . , S_{16}} with data sets S_{1}S_{4}, equilateral triangles; S_{5}S_{8}, squares; S_{9}S_{12}, pentagons; and S_{13}S_{16}; hexagons. Within each subset of distinct polygons, the size of the figures is increasing with the subscript. The perimeter of each polygon S_{k }was divided into 60 equal segments with the segment endpoints ordered clockwise from a fixed initial endpoint. A turtle application was then applied to each polygon to compute the Euclidean distance from each segment endpoint to every other segment endpoint (initial endpoint included). Let x_{S}_{k}^{i }denote the ith endpoint of polygon S_{k}, then the mapping M is defined by
M:S→H^{60}(R),
S_{k}[d_{S}_{k}_{1}d_{S}_{k}_{2}. . . d_{S}_{k}_{60}]
where the columns
d_{S}_{k}_{i}=(d(x_{S}_{k}^{i},x_{S}_{k}^{1}), d(x_{S}_{k}^{i}, x_{S}_{k}^{2}), . . . , d(x_{S}_{k}^{i}, x_{S}_{k}^{60}))^{1}.
The individual column vectors d_{S}_{k}_{j }have intrinsic interest. When plotted as functions of arc length they represent a geometric signal which contains both frequency and spatial information.
The 16, 60×60 distance matrices were input into a publicly distributed version of PROXSCAL. PROXSCAL was run with the following technical specifications: sources—16, objects—60, dimension—4, model—weighted, initial configuration—Torgerson, conditionality—unconditional, transformations—numerical, rate of convergence—0.0, number of iterations—500, and minimum stress—0.0.
In the source space, the different types of polygons are arranged, and hence, classified, along different radii. Magnitudes within each such radial classification indicate polygon size or scale with the smaller polygons located nearer the origin.
The contextual nature of the polygon classification is embodied in the common space configuration. Intuitively, this configuration looks like a single, carefully bent wire loop. When viewed from different angles, as encoded by the source space vectors, this loop of wire looks variously like a triangle, a square, a pentagon, or a hexagon.
The polygons in Example A were regular. In this example, irregular polygons S={S_{1}, . . . , S_{6}} are considered, where S_{1}S_{3 }are triangles and S_{4 } S_{6 }rectangles. The perimeter of each figure S_{k }was divided into 30 equal segments with the preprocessing transformation M computed as in Example A. This produced 6, 30×30 source matrices which were input into PROXSCAL with technical specifications the same as those above except for the number of sources, 6, and objects, 30.
It is notable that two dimensional IDMDS outputs were not sufficient to classify these polygons in the sense that source space separating hyperplanes did not exist in two dimensions.
This example relates to signal processing and demonstrates the analysis tool'"'"'s invariance with respect to phase and frequency modification of time series data. It also demonstrates an entrywise approach to computing the preprocessing transformation M.
The set S={S_{1}, . . . , S_{12}} consisted of sine, square, and sawtooth waveforms. Four versions of each waveform were included, each modified for frequency and phase content. Indices 14 indicate sine, 58 square, and 912 sawtooth frequency and phase modified waveforms. All signals had unit amplitude and were sampled at 32 equal intervals x, for 0≦x≦2π.
Each time series S_{k }was mapped into a symmetric matrix as follows. First, an “empty” nine dimensional, lower triangular matrix T_{k}=(t_{ij}^{k})=T(S_{k}) was created. “Empty” meant that T_{k }had no entries below the diagonal and zeros everywhere else. Nine dimensions were chosen since nine is the smallest positive integer m satisfying the inequality m(m−1)/2≧32 and m(m−1)/2 is the number of entries below the diagonal in an m dimensional matrix. The empty entries in T_{k }were then filled in, from upper left to lower right, column by column, by reading in the time series data from S_{k}. Explicitly: s_{1}^{k}=t_{21}^{k}, the first sample in S_{k }was written in the second row, first column of T_{k}; s_{2}^{k}=t_{31}^{k}, the second sample in S_{k }was written in the third row, first column of T_{k}, and so on. Since there were only 32 signal samples for 36 empty slots in T_{k}, the four remaining entries were designated missing by writing −2 in these positions (These entries are then ignored when calculating the stress). Finally, a hollow symmetric matrix was defined by setting
M(S_{k})=T_{k}+T_{k}^{l}.
This preprocessing produced 12, 9×9 source matrices which were input to PROXSCAL with the following technical specifications: sources—12, objects—9, dimension—8, model—weighted, initial configuration—Torgerson, conditionality—unconditional, transformations—ordinal, approach to ties—secondary, rate of convergence—0.0, number of iterations—500, and minimum stress—0.0. Note that the data, while metric or numeric, was transformed as if it were ordinal or nonmetric. The use of nonmetric IDMDS has been greatly extended in the present embodiment of the tool 100.
The data set S={S_{1}, . . . , S_{9}} in this example consisted of nine sequences with ten elements each; they are shown in Table 1,
The nine source matrices M(S_{k})=(M_{ij}^{k}) were defined by
m_{ij}^{k}=s_{i}^{k}−s_{j}^{k},
the absolute value of the difference of the ith and jth elements in sequence S_{k}. The resulting 10×10 source matrices where input to PROXSCAL configured as follows: sources—9, objects—10, dimension—8, model—weighted, initial configuration—simplex, conditionality—unconditional, transformations—numerical, rate of convergence—0.0, number of iterations—500, and minimum stress—0.0.
This example extends the previous result to demonstrate the applicability of the analysis tool to missing value estimation on noisy, realworld data. The data set consisted of nine categories of bridge data from the National Bridge Inventory (NBI) of the Federal Highway Administration. One of these categories, bridge material (steel or concrete), was removed from the database. The goal was to repopulate this missing category using the technique of the presently preferred embodiment to estimate the missing values.
One hundred bridges were arbitrarily chosen from the NBI. Each bridge defined an eight dimensional vector of data with components the NBI categories. These vectors were preprocessed as in Example D, creating one hundred 8×8 source matrices. The matrices were submitted to PROXSCAL with specifications: sources—100, objects—8, dimension—7, model—weighted, initial configuration—simplex, conditionality—unconditional, transformations—numerical, rate of convergence—0.0, number of iterations—500, and minimum stress—0.00001.
The seven dimensional source space output was partially labeled by bridge material—an application of dyedropping—and analyzed using the following function
where p is an empirically determined negative number, d (x, y) is Euclidean distance on the source space, and X_{A}_{i }is the characteristic function on material set A_{i}, i=1, 2, where A_{i }is steel, A_{2 }concrete. (For the bridge data, no two bridges had the same source space coordinates, hence g_{p }was welldefined.) A bridge was determined to be steel (concrete) if g_{p}(A_{1}, x)>g_{p}(A_{2}, x) (g_{p}(A_{1}, x)<g_{p}(A_{2}, x)). The result was indeterminate in case of equality.
The tool 100 illustrated in
This example demonstrates the use of stress/energy minima to determine network dimensionality from partial network output data. Dimensionality, in this example, means the number of nodes in a network.
89 A fournode network was constructed as follows: generator nodes 1 to 3 were defined by the sine functions, sin(2x), sin(2x+ π/2), and sin(2x+ 4π/3); node 4 was the sum of nodes 1 through 3. The output of node 4 was sampled at 32 equal intervals between 0 and 2π.
The data from node 4 was preprocessed in the manner of Example D: the ijth entry of the source matrix for node 4 was defined to be the absolute value of the difference between the ith and jth samples of the node 4 time series. A second, reference, source matrix was defined using the same preprocessing technique, now applied to thirty two equal interval samples of the function sin(x) for 0≦x≦2π. The resulting 2, 32×32 source matrices were input to PROXSCAL with technical specification: sources—2, objects—32, dimension—1 to 6, model—weighted, initial configuration—simplex, conditionality—conditional, transformations—numerical, rate of convergence—0.0, number of iterations—500, and minimum stress—0.0. The dimension specification had a range of values, 1 to 6. The dimension resulting in the lowest stress/energy is the dimensionality of the underlying network.
Table 2,
From the foregoing, it can be seen that the illustrated embodiment of the present invention provides a method and apparatus for classifying input data. Input data are received and formed into one or more matrices. The matrices are processed using IDMDS to produce a stress/energy value, a rate or change of stress/energy value, a source space and a common space. An output or back end process uses analytical or visual methods to interpret the source space and the common space. The technique in accordance with the present invention therefore avoids limitations associated with statistical pattern recognition techniques, which are limited to detecting only the expected statistical pattern, and syntactical pattern recognition techniques, which cannot perceive beyond the expected structures. Further, the tool in accordance with the present invention is not limited to the fixed structure of neural pattern recognizers. The technique in accordance with the present invention locates patterns in data without interference from preconceptions of models or users about the data. The pattern recognition method in accordance with the present invention uses energy minimization to allow data to selforganize, causing structure to emerge. Furthermore, the technique in accordance with the present invention determines the dimension of dynamical systems from partial data streams measured on those systems through calculation of stress/energy or rate of change of stress/energy across dimensions.
While a particular embodiment of the present invention has been shown and described, modifications may be made. For example, PROXSCAL may be replaced by other IDMDS routines which are commercially available or are proprietary. It is therefore intended in the appended claims to cover all such changes and modifications which fall within the true spirit and scope of the invention.
It is therefore intended that the foregoing detailed description be regarded as illustrative rather than limiting, and that it be understood that it is the following claims, including all equivalents, that are intended to define the spirit and scope of this invention.