IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0626244
(2012-09-25)
|
등록번호 |
US-9547820
(2017-01-17)
|
우선권정보 |
KR-10-2011-0115919 (2011-11-08) |
발명자
/ 주소 |
- Kim, Youn-ho
- Choi, Chang-mok
- Shin, Kun-soo
- Lee, Myoung-ho
- Kim, Jin-kwon
|
출원인 / 주소 |
- Samsung Electronics Co., Ltd.
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
1 인용 특허 :
12 |
초록
▼
A method of classifying an input pattern and a pattern classification apparatus are provided. The method includes enabling an artificial neural network to learn based on learning input data received by an input layer of the artificial neural network, determining classification of an input pattern re
A method of classifying an input pattern and a pattern classification apparatus are provided. The method includes enabling an artificial neural network to learn based on learning input data received by an input layer of the artificial neural network, determining classification of an input pattern received by the input layer of the enabled artificial neural network according to an output value obtained from an output layer of the artificial neural network, the obtained output value being based on the input pattern, updating connection intensities of a plurality of connection lines of the enabled artificial neural network to output a result value indicating the determined classification from the output layer when the input pattern, and determining updated classification of the input pattern according to an updated output value obtained from an output layer of the updated artificial neural network, the obtained updated output value being based on the input pattern.
대표청구항
▼
1. A method of classifying an input pattern, the method comprising: enabling an artificial neural network to learn based on learning input data received by an input layer of the artificial neural network and learning output data outputted by an output layer of the artificial neural network;determini
1. A method of classifying an input pattern, the method comprising: enabling an artificial neural network to learn based on learning input data received by an input layer of the artificial neural network and learning output data outputted by an output layer of the artificial neural network;determining classification of an input pattern received by the input layer according to an output value obtained from the output layer, the obtained output value being based on the input pattern;determining a result value of the determined classification;setting, in response to the determining of the result value, the result value to the output layer so as to form data of the learning output data, and the input pattern to the input layer;updating, in response to the result value being set to the output layer, connection intensities of connection lines of the artificial neural network using the result value set to the output layer when the input pattern is input to the input layer; anddetermining updated classification of the input pattern according to an updated output value obtained from an output layer of the updated artificial neural network, the obtained updated output value being based on the input pattern and the updated connection intensities of the connection lines of the artificial neural network,wherein the result value is a representative value of the determined classification comprising an interval of output values corresponding to the input pattern and including the output value. 2. The method of claim 1, further comprising: outputting learning output data from an output layer of the artificial neural network, the learning output data corresponding to the learning input data. 3. The method of claim 2, wherein the learning output data corresponds to the obtained output value. 4. The method of claim 1, wherein the artificial neural network comprises a single hidden layer. 5. The method of claim 1, wherein the updating of the connection intensities comprises updating a group of the connection lines between a hidden layer and the output layer of the artificial neural network. 6. The method of claim 1, further comprising: determining the connection lines to be updated from among the connection lines of the artificial neural network. 7. The method of claim 6, wherein the determining of the connection lines comprises determining a group of the connection lines between a hidden layer and the output layer. 8. The method of claim 7, wherein the determining of the connection lines comprises determining connection intensities of the group of the connection lines between the hidden layer and the output layer based on a principal component analysis (PCA) method. 9. The method of claim 1, wherein the updating of the connection intensities comprises fixing connection intensities of a group of the connection lines between the input layer and a hidden layer of the artificial neural network, and updating connection intensities of a second group of the connection lines between the hidden layer and the output layer. 10. A non-transitory computer readable recording medium having recorded thereon a program for executing the method of claim 1. 11. A pattern classification apparatus, comprising: memory; anda processor, andwherein the processor includes: a learning unit configured to enable an artificial neural network to learn based on learning input data inputted to an input layer of the artificial neural network and learning output data outputted by an output layer of the artificial neural network;a pattern classification unit configured to determine classification of an input pattern received by the input layer according to an output value obtained from the output layer, the obtained output value being based on the input pattern; anda connection intensity updating unit configured to determine a result value of the determined classification, set, in response to the determination of the result value, the result value to the output layer so as to form data of the learning output data, and the input pattern to the input layer, and update, in response to the result value being set to the output layer, connection intensities of connection lines of the artificial neural network using the result value set to the output layer when the input pattern is input to the input layer,wherein the pattern classification unit is further configured to determine an updated classification of the input pattern according to an updated output value obtained from an output layer of the updated artificial neural network, the obtained updated output value being based on the input pattern and the updated connection intensities of the artificial neural network, andwherein the result value is a representative value of the determined classification comprising an interval of output values corresponding to the input pattern and including the output value. 12. The pattern classification apparatus of claim 11, wherein the learning unit is further configured to output learning output data from an output layer of the artificial neural network, the learning output data corresponding to the learning input data. 13. The pattern classification apparatus of claim 12, wherein the learning output data corresponds to the obtained output value. 14. The pattern classification apparatus of claim 11, wherein the artificial neural network comprises a single hidden layer. 15. The pattern classification apparatus of claim 11, wherein the connection intensity updating unit is further configured to update a group of the connection lines between a hidden layer and the output layer of the artificial neural network. 16. The pattern classification apparatus of claim 11, wherein the connection intensity updating unit is further configured to determine the connection lines to be updated from among the connection lines of the artificial neural network. 17. The pattern classification apparatus of claim 16, wherein the connection intensity updating unit is further configured to determine a group of the connection lines between a hidden layer and the output layer. 18. The pattern classification apparatus of claim 17, wherein the connection intensity updating unit is further configured to determine connection intensities of the group of the connection lines between the hidden layer and the output layer based on a principal component analysis (PCA) method. 19. The pattern classification apparatus of claim 11, wherein the connection intensity updating unit is further configured to fix connection intensities of a group of the connection lines between the input layer and the hidden layer of the enabled artificial neural network, and update connection intensities of a second group of the connection lines between the hidden layer and the output layer.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.