Web-based search system of pattern recognition for the pattern
of industrial component by an innovative technology
Sung-Jung Hsiao
a, Wen-Tsai Sung
b,*, Shih-Ching Ou
b aLearning Science and Technology Laboratory, Department of Computer and Information Science, National Chiao Tung University, Taiwan
b
Bioinformatics & CAD Laboratory, Department of Electrical Engineering, National Central University, 4F, No. 184-1, Kee-Kin 1st Road, An Lo District, 204 Keelung, Taiwan
Received 26 June 2003; accepted 27 June 2003
Abstract
The real-time system uses a recurrent neural network (RNN) with associative memory for training and recognition. This study attempts to use associative memory to apply pattern recognition (PR) technology to the real-time pattern recognition of engineering components in a web-based recognition system with a client–server network structure. Remote engineers can draw the shape of the engineering components using the browser, and the recognition system then searches the component database via the Internet. Component patterns are stored in the database system considered here. Moreover, the data fields of each component pattern contain the properties and specifications of that pattern, except in the case of engineering components. The database system approach significantly improves recognition system capacity. The recognition system examined here employs parallel computing, which increases system recognition rate. The recognition system used in this work is an Internet based client–server network structure. The final phase of the system recognition applies database matching technology to processing recognition, and can solve the problem of spurious states. The system considered here is implemented on the Yang-Fen Automation Electrical Engineering Company as a case study. The experiment is continued for 4 months, and engineers are also used to operating the web-based pattern recognition system. Therefore, the cooperative plan described above is analyzed and discussed here. Finally, these papers propose the tradition methods compare with the innovative methods.
# 2003 Elsevier B.V. All rights reserved.
Keywords: Real-time; Web-based; Pattern recognition; Engineering components; Component database
1. Introduction
Pattern recognition (PR) is widely discussed on the Internet. Weather-forecasting, document analy-sis and recognition, optical character recognition (OCHRE) blocking recognition and prediction, and even financial forecasting all use pattern recognition
[1]. This study considers some recognized procedures with limited capacity. These procedures can be improved but they have not yet yielded an optimal solution to capacity problems involving large quan-tities of data[2–4]. Associative memory is critical in neural networks, and is central to pattern recognition. Many works on pattern recognition have focused on the structure of associative memory [5,6]. The recur-rent neural network (RNN) provides the basis for non-linear associative memory. Significantly, the RNN is very effective in pattern recognition [7,8].
*Corresponding author. Tel.:þ886-2-24325455.
E-mail addresses: song1208@ms5.hinet.net (S.-J. Hsiao), songchen@ms10.hinet.net (W.-T. Sung).
0166-3615/$ – see front matter # 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.compind.2003.06.002
The importance of the Internet is growing, but several local-end pattern recognition programs are still being developed. Consequently, future Internet based pattern recognition methods are likely to focus on integrating science and technology.
The web-based system presented here applies asso-ciative memory technology to component recognition; moreover, it is also a neural network with an RNN structure. Traditional systems adopt RNN for pattern recognition, but only consider character scope recog-nition. In the approach developed here, the system also performs recognition using RNN, but instead consid-ers component pattern. This approach develops a comprehensive web-Based recognition system using an Internet based client–server network structure. Therefore, the database of stored patterns is called the server-end, while the user interface is called the client-end.
The server-end database stores all component warehouse patterns. In many sample patterns, this study proposes using engineering component shape and circuit sign as the basis for sample pattern recognition. Users can input a pattern to search for in the handwriting input area of the client-end. The system launches the recognition task after the search button is clicked. In the recognition process of the training phase, the system uses par-allel computing to improve pattern storage capacity. On the other hand, during the retrieval stage, the web-based system uses database contrast technology to reduce the problem of spurious states produced by the RNN. A simulation is also presented to clarify and corroborate the web-based PR technology. Finally, future developments in the proposed web-based PR framework and algorithm analysis are also discussed.
2. Parallel computing and system analysis This investigation proposes an innovative pattern recognition network for enhancing the network struc-ture of an RNN. In the classical approach[9], an RNN is a discrete-time discretely valued dynamic system characterized by a binary state vector at any given time, t, as follows:
xðtÞ ¼ ½x1ðtÞ; . . . ; xiðtÞ; . . . ; xnðtÞ 2 f1; 1gn (1)
The system behavior is given by the dynamic equa-tion: xiðt þ 1Þ ¼ sgn Xn j¼1 WijXjðtÞ yi " # ; i¼ 1; 2; . . . ; n (2) Point x is fixed for all pattern prototype vectors, x1;x2; . . . ;xp[10]:
xu¼ ½x1ðtÞ; . . . ; xiðtÞ; . . . ; xnðtÞ 2 f1; 1gn (3)
In the approach presented here, x(t) is a record in the pattern database, while x1(t) or xi(t) are fields of any
data record. Moreover, bipolar data take the values of 1 or 1, where 1 and 1 represent points and gaps in the pattern, respectively.
The sample patterns are stored directly in the database system of the server-end via the Internet. A user can modify the database system patterns at any time, and a remote user may establish his or her sample patterns, as illustrated inFig. 1.
In the network parameter, the synaptic matrix, W, and the threshold rector, y, are modifications and improvements of the values in discrete Hopfield net-works [11]. Initially, during the training stage, the records of a pattern database are cut. Next, a parallel computation is employed to determine the W and y values of each segment, and the W and y of each segment are again employed inEq. (2)and, thus the most similar retrieval pattern records in every segment are identified. These similar pattern records are col-lected, and their W and y also are calculated again usingEq. (2). Repeating the computation several times ultimately yields a correct pattern from among numer-ous sample patterns.
The web-based PR system employs a parallel computing architecture [25]. For example, if the
pattern database includes 50 records and the cut number is 10; parallel computation is used to identify the W and y of every group of 10 records. Next, the W and y of every group of 10 records is calculated using Eq. (2). Sets of highly similar pattern records are rearranged into new pattern records. These new pattern records are then collected and their W and y are again calculated, according to the first cut number. The computation is then repeated, until the recognition result is determined, as displayed in
Fig. 2.
The operation of a discrete Hopfield network to manipulate memories involves two phases—storage and retrieval.
2.1. Storage phase
Assume that a set of N-dimensional vectors (binary word), denoted byfxmjm ¼ 1; 2; . . . ; Ng, and is to be
stored. These N vectors are called fundamental memories and represent the patterns to be memorized by the network. Let xm denote the ith element of
the fundamental memory, xm, where the class
m¼ 1; 2; . . . ; N.
From the outer product rule of storage, Hebb’s hypothesis concerning the learning of the synap-tic weight from neuron i to neuron j is generalized as Wji¼ 1 p XN m¼1 xmjxmi (4)
where 1/p is taken as a constant to simplify the mathematical description of information retrieval
[12]. Notably, the learning rule in Eq. (4) is a ‘‘one shot’’ computation.
The normal operation of the Hopfield network is based on the following setting:
Wii¼ 0 for all i; i¼ 1; . . . ; p (5)
Wii¼ 0, preventing positive feedback[13].
Let W denote the P P synaptic weight matrix of the network, with Wjias its jith element.Eqs. (4) and (5)can then be combined into a single equation written in matrix form: W ¼1 p XN m xmxTm N PI (6)
where I represents the P P identity matrix, and W is a symmetric matrix the diagonal line of which has a constant value of zero:
W ¼ W11 W1p .. . } ... Wp1 Wpp 2 6 4 3 7 5 (7) W ¼ 0 W1p .. . 0 ... .. . } ... Wp1 0 2 6 6 6 6 4 3 7 7 7 7 5 (8)
Fig. 2. Parallel computing W and y via the pattern database system.
The threshold of the jth neuron has two modes: yj¼ 0; j¼ 1; . . . ; p (9) or yj¼ XP i¼1 Wij; i¼ 1; . . . ; p (10)
Network memory capacity increases with the thresh-old ofEq. (10) [14].
2.2. Retrieval phase
Given a recognizing pattern vector X as an input, the initial output value is Xð0Þ. Every neuron follow-up output is computed usingEq. (11):
Xjðn þ 1Þ ¼ sgn XP i¼1 WjiXjðnÞ yj ! ¼ sgnðujðnÞ yjÞ ¼ 1 if ujðnÞ > yj xjðnÞ if ujðnÞ ¼ yj 1 if ujðnÞ < yj 8 > < > : (11) Hopfield originally used 0 and 1 as outputs [11]. However, 1 and 1 are now commonly used to allow convenient use of the zero thresholds[15,16].
In Eq. (11), n denotes the number of iterations.
Significantly, the discrete Hopfield network used an asynchronized method to alter individual neuron out-put, and the complete associative memory process employedEq. (12)to describe the chain–state relation-ship:
Xð0Þ ! Xð1Þ ! Xð2Þ ! ! XðkÞ
! Xðk þ 1Þ ! (12) The output remains unchanged by continual iterative computation until the state converges on the stable state. Mathematically, X¼ sgnðWX yÞ, where X is a stable state. Using a synchronization mode to change the network output, changes many results, but the neural network still converges on the stable state simultaneously with the convergence of the partial state on this state. Another partial state can demon-strate a maximum cycle length of two[17].
Although this study uses an asynchronization method to change network output, X converges on the stable state, and sometimes also on the incorrect recall[18]. The X state of final convergence is therefore
used for matching with the original pattern database
[20]. Computing each Hammer distance determines the minimum dH value[19]. Given n pattern records, the Hammer distance is calculated by
dH¼X
P
i¼1
jXi xuij; u¼ 1; 2; . . . ; n (13)
Meanwhile, the minimum value is calculated by dHmin ¼ min X p i¼1 jXi x1ij; Xp i¼1 jXi x2ij; . . . ; Xp i¼1 jXi xnij ( ) (14) If the convergence value of X equals a vector of the sample pattern, xu, then dHmin¼ 0. Conversely, if the
convergent result of the X does not equal a vector of the sample pattern, xu, then dHmin> 0. In the latter
case, X resembles the sample pattern, xu.
3. Establishing and managing the pattern database
In this work, the pattern database is built on a Microsoft SQL Server platform, and the platform
Fig. 3. The recognition system links the database system to the Internet.
integrates the whole recognition system to the Inter-net. The proposed technique uses a web assistant to increase interaction between the system and database, as shown inFig. 3.
Fig. 3displays that the website of the recognition system use Microsoft Internet Information server (IIS). All web pages are located in the IIS, facilitating convenient management by the administrator. Parti-cularly, the administrator can monitor any changes that users make to the contents of the database.
The functioning of the pattern database can be divided into two main parts pattern establishment and data management, both of which can be performed using the browser. When these sample patterns are entered into the pattern database, the approach presented here adopts the web-based and real-time methods. After user input a pattern and clicks the submit button, the pattern is entered into the database. Moreover, simul-taneously with inputting the patterns, users also input their relational properties.
After users input a sample pattern, the administrator can supervise and manage the database with the help of web assistants. This function allows the adminis-trator to view the patterns most recently stored in the database by the browser. Furthermore, the adminis-trator can also modify the database at any time. The whole system is shown inFig. 4.
In the system presented here, web assistants can view the newest data in the pattern database. More-over, users can also view the field data of pattern number and pattern builder, as illustrated inFig. 5. The pattern database designed here employs a relational mode to establish the data tables for viewing by users. The pattern registrar and pattern data simultaneously
are separated by the relational mode. This technique simplifies the pattern database. Fig. 6 presents the relational graph of the data table.
This investigation establishes the data table using the method of ‘‘one to many’’ to create the data table. Once the recognition is accomplished, the recognition system identifies the correct pattern, and simulta-neously users can also view the pattern properties.
4. Storage capacity analysis and improvement As an important model of associative memory, the Hopfield network has been comprehensively researched and applied to pattern recognition using the sum-of-outer products [11]. Additional research has examined the asymmetric or generalized Hop-field model using other learning algorithms, since the memory capacity of the Hopfield network using the sum-of-outer products scheme is extremely low
[21–23].
Hopfield was the first to determine the number of stable patterns for the Hopfield RNN at 0.15P (for P neurons) [11]. Since then, many other studies have obtained results that indicate better performance cap-ability.
The capacity of a Hopfield RNN is its number of stable states, C. Obviously, C depends on the weight matrix, which is taken to be symmetric with the values on the diagonal constantly being zero. McEliece et a1.
[21]demonstrated that P
½4lnðPÞ< C < P
½2lnðPÞ (15)
Fig. 5. The system homepage shows the component pattern data and builder.
For example, for 100 neurons, C satisfies 5 < C< 10, where C denotes the number of data records in the stable state. The memory capacity of a discrete Hopfield network has an upper limit. Given P neurons,
Eq. (15)yields: Mmax¼
P
2lnðPÞ; (16)
where Mmaxis the maximum memory capacity.
Amit[24]stated that the number of neurons, P, is 99% correct in the retrieval phase, and the number of the stored data records is limited by the following formula:
M P
4 lnðPÞ; (17)
where M is the memory capacity.
Notably, M inEqs. (16) and (17)becomes the basis of the divided segment in the pattern database, being the number used for distributed computation.
Writing a component pattern on a computer requires P¼ 240ð15 16 matrix) neurons and has a P2 P ¼ 57,360 weight value for the recollection.
Therefore, P¼ 240 in Eqs. (16) and (17), and the different capacities are determined based on the num-ber of cut recorded patterns in the database. Assume thatEq. (17)is used to determine the cut number with 100% recognition: M P 4lnðPÞ; M 240 4lnð240Þ; ðP ¼ 240Þ; M 10 ðrecordsÞ (18) Accordingly, the partition number of the pattern data-base was set to ten. Consequently, the pattern for each group of ten records was treated as a segment, with the W and y values of each segment being calculated separately.
5. Implementing the web-based pattern recognition system
This section further considers the implementation of the proposed pattern recognition system. The pat-tern database system was established first at the server-end, and Microsoft SQL Server was used as the data management platform. Input was a dynamic action for sample patterns, and a real-time, web-based method
was used for pattern input. Notably, the new learning patterns can be built at any time. Once established, the sample patterns can be updated, modified, and deleted. Namely, the above mentions fully conform to the rules which build the pattern database, for which refer to
Fig. 7.
Using the method displayed inFig. 2and distributed computation of partition database efficiently solves the capacity problem. The new learning pattern used a dynamic method and thus can perform pattern recog-nition at any time. The measured pattern database and client-end operation results are recorded on a web page, as displayed inFig. 8.
InFig. 8, the right of the web page is the client-end and the left of the web page is the server-end. If the user inputs a pattern at the client-end, the result recognized as correct is displayed at the server-end even when the source pattern of the client-end suffers
Fig. 7. Using a web-based method to build a pattern database (prototype database).
Fig. 8. Patterns are inputted at the client-end and displayed at the server-end.
from noise interference. The proposed recognition system has already overcome many of the problems of previous systems. For example, the proposed sys-tem has substantially better capacity and accuracy than existing systems, and the neural network with distributed computation is highly efficient.
This investigation now analyzes the convergence of the recognition system, with reference to Lippmann’s experiment in which the inputs were assumed to take the valueþ1 for black points and 1 for white points. A selected pattern is distorted by randomly and independently reversing each point of the pattern from þ1 to 1 and vice versa, with a probability of 0.25, and then testing the network using the corrupted pattern. Fig. 9 illustrates the recognition results for a component pattern. The patterns produced by the network after 30, 60, 100, 150, 200, and 238 iterations reveal a steady increase the resemblance of the net-work output to the component pattern. Indeed, after 238 iterations, the network converges onto the correct form of the component pattern.
Fig. 9shows the correct pattern of stable
conver-gence after 238 iterations at the server-end.
Next, for a case of spurious states, the noisy pattern is inputted at the client-end, and the partial recall
pattern is recognized at the server-end. The pattern is not recalled correctly because a matching sample pattern was not inputted, as presented inFig. 10.
The system increased the matching of the pattern database. Accordingly, the accuracy displays no increase in Fig. 10. When the recognition result matched the pattern database, it converged to an accurate pattern, as inFig. 11.
Fig. 9. Complete system of pattern recognition in the convergent process.
6. Cooperative example
Using the web-based recognition system presented here, our laboratory and Yang-Fen Automation Elec-trical Engineering Company implemented a techno-logical co-operation plan during January, 2002. Yang-Fen Automation Electrical Engineering Company installs industrial power distribution equipment. Stock levels vary for each power distribution component. Furthermore, Yang-Fen is involved in projects all over the world. Formerly, engineers would ask head office
about stock levels of these components via the tele-phone, meaning frequent communication mistakes occurred. More recently however Yang-Fen estab-lished an on-line method of performing these query tasks. Unfortunately, some engineers are prone to forget the name of certain engineering components, causing further problems in queries regarding stock levels.
Next, our laboratory cooperated with Yang-Fen Company on an experiment using real-time methods of searching for component database patterns via the Internet, where the search itself is a recognized task. First, shape of each component is entered into the component database. Each component uses their shape to become the pattern of component database. The component database joins the specifications in the other field when these component database patterns are established, as illustrated in Fig. 12.
Fig. 11. Conditions for accurate convergence existed when the recognition result matched the pattern database.
Fig. 12. Relational component pattern database.
These component patterns are shown in the follow-ing figures from the web page of Yang-Fen Company, for example,Fig. 13.
Each component displays a pattern, and some com-ponents share the same pattern. The following figure displays an inventory of components, such asTable 1. System users can log on to the server-end compo-nent recognition homepage without being restricted by spatial constraints. The user can then input a self-drawn component pattern, and the system will attempt to identify the pattern after recognizing button is clicked (Table 2).
From the recognition statistics of Yang-Fen Company from January to April 2002, their engineers were not familiar with the operation of web-based recognition system in January, and consequently the
recognition rate was low. However, the recognition rate improved in February as the engineers became familiar with the system. Database modification was conducted through a cooperative process of modifying original component patterns. Importantly component
Table 1
Two different components that share the same recognition pattern Pattern Pattern name Actual pattern
Pilot lamp (circular form)
Pilot lamp (circular head)
Table 2
Recognized statistics for cooperative experiment Month Total recognition times Correct recognition times Incorrect recognition times Recognition ratio (%) January 232 163 69 70.26 February 256 228 28 89.06 March 247 230 17 93.12 April 269 262 7 97.40
Fig. 14. Analysis of cooperative plan based on history.
patterns were not allowed to be too similar to one another, to increase the recognition rate of the system. Next, the data that each engineer entered into the recognition system for the times of success and failure are listed from January to April (Figs. 14 and 15).
7. Traditional methods compare with innovative methods
These engineers that used traditional method
[26–29] of search cannot use the search of remote
component database by the Internet until they must memorize each name or number of component of power distribution.
There are some obvious disadvantages in the tradi-tional method; we must pay attention to them. If these engineers were not familiar with those data which will be searched, they should not process the search of remote database. Simultaneously, it will take much time to look up the data books of component, and the completed task will be delayed. On the other hand, these engineers’ background of education will also influences the quality of search easily.
According to these disadvantages which used tradi-tional way, our study presented the innovative search system of pattern recognition for the pattern of indus-trial component. Remote engineers can draw these
figures of industrial component directly in the brow-ser, and after search they will get these relational inventories of industrial component in the real-time situation.
In our paper, we compared the traditional method with the innovative method, and the result was pre-sented in Fig. 16.
In our compare, the traditional methods only input the name or number of industrial component, and they have not used the technology of pattern recognition yet. Consequently, the contrived factor will influence the traditional methods easily. In our approach, we use the technology of artificial intelligence to do the work of pattern recognition, and these engineers will have not any heavy burden on their tasks.
According to the statistics for the average of each hundred search from January to April, the traditional methods cannot control the correct rate and time of system recognition more effectively. Using the inno-vative method to do the search work, these engineers are becoming more and more proficient in the operat-ing skill of search.
8. Algorithm analyses
The proposed algorithm was based on the theory of Lippmann[16], but with improvements. Notably, the
newly proposed our approaches were included. Our approach can be easily implemented using a computer program.
The algorithm uses the following steps to identify the correct pattern in the pattern database.
Step 1. Calculate the number of cut records of the pattern database:
M¼ P 4lnðPÞ
ðP is the total number of neuronsÞ
(19)
Step 2. Every set of M records comprise a segment from among the records of the pattern database. More-over, the entire database is divided into segments, as follows M, 2M, 3M, . . ., (the maximum number of cut records¼ CRmax)M, and the W and y values are
computed:
Step 3. In the retrieval stage, n denotes the number of iterations, and X indicates that the pattern will be recognized:
Step 4. Every convergent X value in Step 3 is deter-mined, and the matching pattern database determines the minimum Hamming distance:
WðMÞ¼ 1 p XM K¼1 xKxTK M PI; yjðMÞ ¼ XP i¼1 WjiðMÞ; i¼ 1; . . . ; P Wð2MÞ¼ 1 p X2M K¼Mþ1 xKxTK M PI; yjðMÞ ¼ XP i¼1 Wjið2MÞ; i¼ 1; . . . ; P Wð3MÞ¼ 1 p X3M K¼2Mþ1 xKx T K M PI; yjð3MÞ¼ XP i¼1 Wjið3MÞ; i¼ 1; . . . ; P .. . WðCRmaxMÞ¼ 1 p X CRmaxM K¼CRmaxMþ1 xKxTK M PI; yjðCRmaxMÞ¼ XP i¼1 WjiðCRmaxMÞ; i¼ 1; . . . ; P (20) XjðMÞðn þ 1Þ ¼ sgn X P i¼1 WjiðMÞXiðnÞ yjðMÞ ! Xjð2MÞðn þ 1Þ ¼ sgn X P i¼1 Wjið2MÞXiðnÞ yjð2MÞ ! Xjð3MÞðn þ 1Þ ¼ sgn XP i¼1 Wjið3MÞXiðnÞ yjð3MÞ ! .. . XjðCRmaxMÞðn þ 1Þ ¼ sgn X P i¼1
WjiðCRmaxMÞXiðnÞ yjðCRmaxMÞ
! (21) dHminðMÞ¼ min XP i¼1 jXijðMÞðnþ1Þ x1ij; XP i¼1 jXðnþ1ÞijðMÞ x2ij; . . . ; XP i¼1 jXijðMÞðnþ1Þ xMi j ( ) dHminð2MÞ¼ min XP i¼1 jXðnþ1Þ ijð2MÞ x Mþ1 i j; XP i¼1 jXðnþ1Þijð2MÞ xMþ2i j; . . . ; XP i¼1 jXijð2MÞðnþ1Þ x2Mi j ( ) dHminð3MÞ¼ min XP i¼1 jXðnþ1Þ ijð3MÞ x 2Mþ1 i j; XP i¼1 jXðnþ1Þijð3MÞ x2Mþ2i j; . . . ; XP i¼1 jXijð3MÞðnþ1Þ x3Mi j ( ) .. .
dHminðCRmaxMÞ¼ min
XP i¼1 jXðnþ1Þ ijðCRmaxMÞ x ðCRmax 1ÞMþ1 i j; XP i¼1 jXijðCRðnþ1Þ maxMÞ x ðCRmaxM 1ÞMþ2 i j; . . . ; XP i¼1 jXijð3MÞðnþ1Þ xCRmaxM i j ( ) (22)
Step 5. dHmin, determined in Step 4 can specify that
the X is the most similar to the x (sample patterns). These patterns are combined as new pattern records. Step 2 is revisited and repeated until the output of Step 5 equals one.
Step 6. Finally, the sample pattern, x, identified as a correctly recognized pattern.
The recognition method presented here is new, and easily can be used to write a web page with a pattern recognition function.
9. Conclusions and future work
The application of component recognition on the Internet is still immature. This work applies a real-time, web-based method to recognize network pat-terns on the Internet. The development of a pattern database can solve numerous problems in recognition technology. This work has reached some new solu-tions of improved recognition, as follows:
1. Using the matching technology of a pattern database to determine the patterns that are most similar to one another reduces spurious states of RNN and increases neural network recognition rate.
2. Utilizing the pattern database to establish a learning pattern, and overcoming the problem of limited RNN capacity.
3. The web-based approach uses Internet, and any user can use a browser to connect to the server-end via the Internet. Furthermore, users can recognize the source pattern immediately after inputting the training pattern.
Numerous recognition programs must be run on local machines, and these programs have limited compatibility with many common operating systems. However, transplanting these programs to the Internet can cause some difficulties in Common Gateway Interface (CGI).The program presented here is built in a web-server environment. Notably, the perfor-mance of the program is characterized by a lack of delay because the system performs learning and recognition in real-time.
The proposed recognition system is managed using the back-end database system. After the user logs on to
the system, all patterned data is stored in the database. This approach is new, and guarantees the complete-ness and security of the patterned data.
With further development, the proposed recogni-tion system will be able to be widely applied to electronic commerce (EC). For example, if the ser-ver-end was a bank, an autograph (as a sample pattern) could be remotely registered in a home or office. The signature pattern would then be recog-nized at the server-end. Consequently, the individuals using the network would benefit from more secure transactions, and electronic commerce would be further promoted.
Acknowledgements
The authors would like to thank the National Chiao Tung University of Taiwan and National Central University of Taiwan for financially supporting this research.
References
[1] S. Singh, A long memory pattern modeling and recognition system for financial forecasting, Pattern Analysis and Applications 2 (3) (1999) 264–273.
[2] S. Kak, Better web searches and prediction with instanta-neously trained neural networks, IEEE Intelligent Systems 14 (6) (1999) 78–81.
[3] R.P.W. Duin, Superlearning and neural network magic, Pattern Recognition Letters 15 (1994) 215–217.
[4] M.A. Kraaijveld, R.P.W. Duin, The effective capacity of multilayer feedforward network classifiers, in: Proceedings of the 12th International Conference on Pattern Recognition (ICPR 94), Israel, vol. B, 1994, pp. 99–103.
[5] Z. Tan, M.K. Ali, Pattern recognition with stochastic resonance in a generic neural network, International Journal of Modern Physics C 11 (8) (2000) 1585–1593.
[6] M. Perus, Neural networks as a basis for quantum associative networks, Neural Network World 10 (6) (2000) 1001–1013.
[7] R.K. Brouwer, An integer recurrent artificial neural network for classifying feature vectors, International Journal of Pattern Recognition and Artificial Intelligence 14 (3) (2000) 335–339. [8] R.K. Brouwer, A fuzzy recurrent artificial neural network for pattern classification, International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems 8 (5) (2000) 525– 538.
[9] Y. Kamp, M. Hasler, Recursive Neural Networks for Associative Memory, Series in Systems and Optimization, Wiley/Interscience, England, 1990, pp. 10–34.
[10] V. Gimenez, L. Aslanyan, J. Catellanos, V. Ryazanov, Distribution functions as attractor for recurrent neural net-works, Pattern Recognition and Image Analysis 11 (3) (2001) 492–497.
[11] J.J. Hopfield, Neural networks and physical systems with emergent collective computational abilities, Proceedings of the National Academy of Sciences of the United States of America 79 (1982) 2554–2558.
[12] S. Haykin, Neural Networks: A Comprehensive Foundation, vol. 2n, Macmillan, New York, 1999.
[13] J.J. Hopfield, D.W. Tank, Computing with neural circuits: a model, Science 233 (1986) 625–633.
[14] B. Mueller, J. Reinhardt, M.T. Strickland, Neural Networks, Springer-Verlag, Berlin, Heidelberg, 1995.
[15] J.M. Zurada, Artificial Neural Systems, West Publishing, St. Paul, UN, 1992.
[16] R.P. Lippmann, An Introduction to Computing with Neural Nets, IEEE ASSP Magazine, 1987, pp. 4–22, also reprinted in neural networks: C. Lau (Ed.), Theoretical Foundations and Analysis, IEEE Press, New York, 1992, pp. 5–23.
[17] W.A. Little, G.L. Shaw, Analytical study of the memory storage capacity of a neural network, Mathematical Bios-ciences 39 (1) (1978) 281–290.
[18] S. Haykin, Neural Networks: A Comprehensive Foundation, Macmillan, New York, 1994.
[19] J. Ma, A neural network approach to real-time pattern recognition, International Journal of Pattern Recognition and Artificial Intelligence 15 (6) (2001) 934–947.
[20] J.W. Ma, The stability of the generalized Hopfield networks in randomly asynchronous mode, Neural Networks 10 (6) (1997) 1109–1116.
[21] R.E. McEliece, E.C. Posner, E.R. Rodernich, S.S. VenKa-tesh, The capacity of the Hopfield associative memory, IEEE Transactions on Information Theory 33 (2) (1987) 461–483.
[22] L.F. Abbott, T.B. Kepler, Optimal learning in neural network memories, Journal of Physics A: Mathematical General 22 (1989) 711–717.
[23] S.S. Venkatesh, D. Pitts, Linear and logarithmic capacities in associative memory, IEEE Transactions on Information Theory 35 (1989) 558–568.
[24] D.J. Amit, Modeling Brain Function: The World of Attractor Neural Networks, Cambridge University Press, New York, 1989.
[25] D.E. Culler, J.P. Singh, A. Gupta, Parallel Computing Archit-ecture: A Hardware/Software Approach. Morgan Kaufmann, Los Altos, CA, 1999.
[26] K.N. Gudivada, V.V. Raghavan, W.I. Grosky, R. Kasanagottu, Information retrieval on the World Wide Web, IEEE Internet Computing 1 (5) (1997) 58–68.
[27] S. Lawrence, C.L. Giles, Searching the World Wide Web, Science 280 (1998) 98–100.
[28] S. Lawrence, C.L. Giles, Searching the web: general and scientific information access, IEEE Communication 37 (1) (1999) 116–122.
[29] H. Lu, L. Feng, Integrating database and World Wide Web technologies, World Wide Web 1 (2) (1998) 73–86.
Sung-Jung Hsiao was born in Kee-lung, Taiwan in 1969. He received the BS degree in electrical engineering from the National Taipei University of Technology, Taiwan, in 1996 and the MS degree in computer science and information engineering from National Central University, Taiwan in 2003. Currently, he is studying PhD at the Department of Computer and Informa-tion Science, NaInforma-tional Chiao Tung University as a researcher of Learning Science and Technology Laboratory. He has had the work experience of research and design at the famous computer company of Acer Universal Computer Co., Mitsubishi, and FIC. He has presented some papers at the international conference and journal. His research interests include the learning science and technology, pattern recognition, image processing, CAD/VR graphics, and artificial intelligence.
Wen-Tsai Sung is a PhD candidate at Department of Electrical Engineering, National Central University in Taiwan. His research interests include computer aided design, web-based learning system, bioinformatics and virtual reality. He has published a number of international journal and conferences papers related to these areas. He received a BS degree from the Department of Industrial Educa-tion, National Taiwan Normal University, Taiwan in 1993 and received a MS degree from the Department of Electrical Engineering, National Central University, Taiwan in 2000. He has win the dragon thesis award; master degree thesis be recognized the most outstanding academic research. The thesis entitle is: ‘‘Integrated computer graphics system in a virtual enviro-nment’’. Sponsor is Acer Foundation (Acer Universal Computer Co.). Currently, he is studying PhD at the Department of Electrical Engineering, National Central University as a researcher of Bioinformatics & CAD Laboratory.
Shih-Ching Ou is working with the Department of Electrical Engineering, National Central University as a senior professor. His research interests include computer aided design, e-learning sys-tem, and virtual reality, etc. He has published a number of international journal and conferences papers related to these areas. Currently, he is the chief of Bioinformatics & CAD Laboratory.
