US5768422A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method for training an adaptive statistical classifier to discriminate against inproper patterns
Assignee: Apple Inc
•
Filing: 1995-08-08
•
Granted: 1998-06-16
•
Expires: 2015-08-08
Abstract
Abstract A statistical classifier that can be used for pattern recognition is trained to recognize negative, or improper patterns as well as proper patterns that are positively associated with desired output classes. A set of training samples includes both the negative and positive patterns, and target output values for the negative patterns are set so that no recognized class is indicated. The negative patterns are selected for training with less frequency than the positive patterns, and their effect on training is also modified, so that training is focused more heavily on positive patterns.
Inventors
- 👤 Larry S. Yaeger
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-26 16:07
CPC Classifications
G; G06; G06F; G06F18/00; G06F18/20; G06F18/21
Claims Analysis
Total Claims: 15
Claim 1
Independent Claim
210 words
A method for training a statistical classifier to estimate the probabilities that input patterns are associated with each of a predetermined set of classes, comprising the steps of: selecting a first, positive set of training patterns each associated with a class in said set; selecting a second, negative set of training patterns not associated with any class in said set; combining said first set and said second sets into a training set; processing training patterns in said training set through a process comprising the following steps for each training pattern that is processed; computing a set of target values corresponding to each class in said set of classes, such that: for a training pattern from said first, positive set, the target value corresponding to its associated class is substantially equal to a first predetermined value, and the other target values of said set of target values are all substantially equal to a second predetermined value that is substantially different from said first predetermined value, and; for a training pattern from said second set, negative set, all of said target values are substantially equal to said second predetermined value; and providing said training pattern and said set of target values to a statistical classifier and training said classifier in accordance therewith.
Claim 2
Independent Claim
210 words
A method for training a statistical classifier to estimate the probabilities that input patterns are associated with each of a predetermined set of classes, comprising the steps of: selecting a first, positive set of training patterns each associated with a class in said set; selecting a second, negative set of training patterns not associated with any class in said set; combining said first set and said second sets into a training set; processing training patterns in said training set through a process comprising the following steps for each training pattern that is processed; computing a set of target values corresponding to each class in said set of classes, such that: for a training pattern from said first, positive set, the target value corresponding to its associated class is substantially equal to a first predetermined value, and the other target values of said set of target values are all substantially equal to a second predetermined value that is substantially different from said first predetermined value, and; for a training pattern from said second set, negative set, all of said target values are substantially equal to said second predetermined value; and providing said training pattern and said set of target values to a statistical classifier and training said classifier in accordance therewith.
Claim 3
Dependent Claim
37 words
The method of claim 1 wherein said statistical classifier comprises a portion of a recognition system that processes correct and incorrect segment patterns and produces a recognition interpretation associated with a corresponding subset of said segment patterns.
Claim 4
Dependent Claim
43 words
The method of claim 2 wherein said recognition system is a system for recognizing handwritten words, said correct and incorrect segment patterns comprise groups of strokes, said statistical classifier is trained to classify characters, and said recognition interpretation is a sequence of characters.
Claim 5
Dependent Claim
13 words
The method of claim 3 wherein said correct segment patterns correspond to characters.
Claim 6
Dependent Claim
54 words
The method of claim 1 wherein the step of processing training patterns includes the steps of assigning a probability of usage factor to said training patterns randomly selecting individual training patterns and selectively providing the selected training patterns to the classifier or skipping the selected patterns in accordance with said probability of usage factor.
Claim 7
Dependent Claim
28 words
The method of claim 5 wherein the training patterns in said first set have a different probability of usage factor from the training patterns in said second set.
Claim 8
Dependent Claim
35 words
The method of claim 6 wherein the training patterns in said first set have a probability of usage factor which is higher than the probability of usage factor for training patterns in the second set.
Claim 9
Dependent Claim
63 words
The method of claim 5 further including the steps of detecting whether the classifier produces a probability estimate which correctly identifies the class associated with a training pattern from said first set, and assigning a probability of usage factor to correctly identified training patterns which is different from a probability of usage factor assigned to incorrectly identified training patterns from said first set.
Claim 10
Dependent Claim
31 words
The method of claim 8 wherein the probability of usage factor assigned to correctly identified training patterns is less than the probability of usage factor assigned to incorrectly identified training patterns.
Claim 11
Dependent Claim
31 words
The method of claim 8 further including the steps of repeatedly processing training patterns in the classifier over multiple successive iterations, and increasing the probability of usage factor for later iterations.
Claim 12
Independent Claim
176 words
A method for training a statistical classifier to estimate the probability that an input pattern belongs to a predetermined class, comprising the steps of: defining a first, positive set of training patterns, each of which is labeled as belonging to a respective one of a plurality of predetermined classes; defining a second, negative set of training patterns which are not associated with any of said predetermined classes; processing said training patterns in the statistical classifier to produce output values and associating a target value with each training pattern, wherein the target value associated with a training pattern from said first, positive set designates the class to which said pattern belongs, and the target value associated with a training pattern from said second, negative set designates none of said classes; determining an error value for each processed pattern, based upon the difference between the target value associated with the pattern and the output value produced by the classifier in response to the processed pattern; and adjusting operating parameters of said classifier in accordance with said error value.
Claim 13
Independent Claim
176 words
A method for training a statistical classifier to estimate the probability that an input pattern belongs to a predetermined class, comprising the steps of: defining a first, positive set of training patterns, each of which is labeled as belonging to a respective one of a plurality of predetermined classes; defining a second, negative set of training patterns which are not associated with any of said predetermined classes; processing said training patterns in the statistical classifier to produce output values and associating a target value with each training pattern, wherein the target value associated with a training pattern from said first, positive set designates the class to which said pattern belongs, and the target value associated with a training pattern from said second, negative set designates none of said classes; determining an error value for each processed pattern, based upon the difference between the target value associated with the pattern and the output value produced by the classifier in response to the processed pattern; and adjusting operating parameters of said classifier in accordance with said error value.
Claim 14
Dependent Claim
34 words
The method of claim 11 further including the step of modifying the error values associated with training patterns from said second set, relative to error values associated with training patterns from said first set.
Claim 15
Dependent Claim
31 words
The method of claim 12 wherein the error values associated with training patterns from said second set are reduced relative to error values associated with training patterns from said first set.
References Cited (Backward Citations)
US5161204A
Cited
US5577166A
Cited
US5337371A
Cited
US5204914A
Cited
US5555317A
Cited
US5392363A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2001-06-15 | 2001-12-15 | 2002-06-16 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2005-06-15 | 2005-12-15 | 2006-06-16 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2009-06-15 | 2009-12-15 | 2010-06-16 | $7700.00 | Window Closed / Lapsed |