associative pattern memory (apm) larry werth july 14, 2007
TRANSCRIPT
![Page 1: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/1.jpg)
Associative Pattern Memory (APM)Larry WerthJuly 14, 2007
![Page 2: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/2.jpg)
Introduction and Background of APM
• Human Associative Pattern Memory• Computer Implemented APM• Basis for Two Successful Startup Companies• Six Patents Granted and Others Pending• Successful Implementation of NKS
![Page 3: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/3.jpg)
Objective of My Presentation
• Describe the APM Concept & Implementation• Describe its Advantages / Features• Identify Types of Applications• Describe its Current Status and Future Goals
![Page 4: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/4.jpg)
Origin of Concept
• Randomly Connected Neural Network Models• States Sequence Terminates in a Cycle• Randomly Map Each State to an Input Pattern• Sampled Pattern Value & Current State
Determine Next State• The Ultimate Cycle Represents the Input
Pattern• Cycles Form the Basis of the APM
![Page 5: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/5.jpg)
Cycle PropertiesRandomly Connected DFA’s
Expected# Expected# Expected#Fraction
Terminal Number TransitionTerminal
Total States (N) States(S) Cycles(C) States(T) States(F)100 12 3 7 .121,000 40 4 20 .04010,000 125 5 63
.0125100,000 396 6 198
.003961,000,000 1,253 8 627
.00125310,000,000 3,963 9 1982
.000396100,000,000 12,533 10 6267
.00012531,000,000,000 39,632 11 19817 .0000396
![Page 6: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/6.jpg)
Conceptual Implementation of APM
State Array
Next State Array (Value = 0)
Next State Array (Value = 1)
Input Pattern Array
Pattern Address
Pattern Value
Next State Address
CurrentState Address
Response Array
Respond to Pattern(Read From Cycle Addresses)
Train Pattern(Write to Cycle Addresses)
State Array: Filled with Random Pattern AddressesNext State Arrays : Filled with Random State AddressesResponse Array: Assigned Responses to Patterns
![Page 7: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/7.jpg)
Solution to Multiple Cycles
• Introduce a Refractory Period• A State Can Not Occur Again Until After a
Specified Number of Steps• Establishes a Minimum Cycle Length• Assures One Cycle Per Input Pattern
Independent of Initial State• Input Pattern is Represented By a Single
Sequence of Random Addresses in Memory
![Page 8: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/8.jpg)
Minimum Cycle Length Example
• Number of States: 1,000,000• Minimum Cycle Length: 3,700• Probability of a Second Cycle of 3,700 in
Length: 1 in 1,000,000
Based on the probability of not picking one of 3700 in 1,000,000 after 3700 tries.
![Page 9: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/9.jpg)
Response/Recognition Capacity
• During Training Desired Responses are Written to Cycle Addresses in Response Memory
• Problem: Response Memory Fills UP Quickly• Any Cycle Address has Memory of Previous
Input Sample Values• Do Not Need to Use All Cycle Addresses• Solution: Vertical Sensors
![Page 10: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/10.jpg)
Vertical Sensor Cycle Detection
State Addresses Plane With Cycle
Vertical Sensors Detect Presence Absence of Cycle
Upper Memory Plane forms New Input Pattern Based on Sensor Status
![Page 11: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/11.jpg)
Vertical Sensor Implementation
• Number of States: 1,000,000• Minimum Cycle Length: 3,700• One of 270 Addresses are in Cycles• Vertical Sensor Field Size: 135• Probability Field Contains Cycle Address: .5• Vertical Sensor Determines Bit Status of Hash
Values that Addresses Response Memory
![Page 12: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/12.jpg)
Fuzzy Hash
• Similar Input Patterns Produce Similar Cycles• Similar Input Patterns Generate the same or
Similar Hash Codes• Multiple Independent Hash Codes are
Generated By One Cycle (One Input Pattern)• A Voting Mode For Response Identification
Contributes to Fuzzy Recognition
![Page 13: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/13.jpg)
Advantages of Using Cycles
• Creates a Fuzzy Hash• Simple and Fast Implementation• Common Language for Different Pattern Types • Spatial and Temporal Integration to Form New
Higher Level Input Patterns• Automatic Segmentation of Time Varying
Patterns
![Page 14: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/14.jpg)
Applications
• Actual Applications: Hand Printed Character Recognition, Machine Vision, Video Compression, Financial Pattern Forecasting
• Signal Processing – Vector Quantization• Video Surveillance – Smart Cameras• Video Object Tracking• Stereo Vision
![Page 15: Associative Pattern Memory (APM) Larry Werth July 14, 2007](https://reader036.vdocuments.site/reader036/viewer/2022082611/56649e5c5503460f94b54422/html5/thumbnails/15.jpg)
Current Status and Objectives
• Software Library Written in C/C++• Objective: General Purpose Tool for Pattern
Recognition Development• Looking for a Business Partner• Software Will be Available on Our Web Site
www.netwerth.net