22

Quality Improvement using Data Mining in

Manufacturing Processes
Shu-guang He1, Zhen He1, G. Alan Wang2 and Li Li3
1School
of Management, Tianjin University, Tianjin,
2Pamplin College of Business, Virginia Polytechnic Institute and State University,
3School of Electronics and Information Engineering, Tianjin Professional College, Tianjin,
1,3P.R.China
2U.S.A.

1. Introduction
Nowadays, manufacturing enterprises have to stay competitive in order to survive the
competition in the global market. Quality, cost and cycle time are considered as decisive
factors when a manufacturing enterprise competes against its peers. Among them, quality is
viewed as the more critical for getting long-term competitive advantages. The development
of information technology and sensor technology has enabled large-scale data collection
when monitoring the manufacturing processes. Those data could be potentially useful when
learning patterns and knowledge for the purpose of quality improvement in manufacturing
processes. However, due to the large amount of data, it can be difficult to discover the
knowledge hidden in the data without proper tools.
Data mining provides a set of techniques to study patterns in data “that can be sought
Open Access Database www.intechweb.org

automatically, identified, validated, and used for prediction” (Witten and Frank 2005).
Typical data mining techniques include clustering, association rule mining, classification,
and regression. In recent years data mining began to be applied to quality diagnosis and
quality improvement in complicated manufacturing processes, such as semiconductor
manufacturing and steel making. It has become an emerging topic in the field of quality
engineering. Andrew Kusiak (2001) used a decision tree algorithm to identify the cause of
soldering defects on circuit board. The rules derived from the decision tree greatly
simplified the process of quality diagnosis. Shao-Chuang Hsu (2007) and Chen-Fu Chien
(2006 and 2007) demonstrated the use of data mining on semiconductor yield improvement.
Data mining has also been applied to product development process (Bakesh Menon, 2004)
and assembly lines (Sébastien Gebus,2007). Some researchers combined data mining and
traditional statistical methods and applied to quality improvement. Examples are the use of
MSPC (multivariate statistical control charts) and neural networks in detergent-making
company (Seyed Taghi Akhavan Niaki, 2005; Tai-Yue Wang, 2002), the combination of
automated decision system and six sigma in the General Electric financial Assurance
businesses (Angie Patterson, 2005), the combined used of decision tree and SPC with data
from Holmes and Mergen (Ruey-Shiang Guh, 2008), the use of SVR (support vector
regression) and control charts (Ben Khediri ISSam, 2008), the use of ANN (artificial neural
Source: Data Mining and Knowledge Discovery in Real Life Applications, Book edited by: Julio Ponce and Adem Karahoca,
ISBN 978-3-902613-53-0, pp. 438, February 2009, I-Tech, Vienna, Austria

www.intechopen.com
358 Data Mining and Knowledge Discovery in Real Life Applications

network), SA ( simulated annealing) and Taguchi experiment design (Hsu-Hwa Chang,

2008). Giovanni C Porzio(2003) has presented a methods for visually mining off-line data
with combination of ANN and T2 control chart and to identify the assignable variation
automatically.
Although techniques are available to learn numerous patterns and knowledge hidden in
mass data, it can be difficult to identify those patterns that can be directly applicable. Kaidi
Zhao (2005) developed a visualization tool, named as Opportunity Map, which can help
identify useful and actionable knowledge quickly. Mu-Chen Chen (2007) proposed a
method that ranks the rules learned by data mining techniques using DEA (Data
Envelopment Analysis).
This work focuses on using data mining for quality improvement in manufacturing
processes. The chapter is organized as follows. In section 2, we introduce a knowledge-
based continuous quality improvement model in comparison with DMAIC for six-sigma. In
section 3, we explain parameter optimization, quality diagnosis and service data analysis for
quality improvement. In section 4, we propose a system framework for quality
improvement in manufacturing processes. Finally, section 5 concludes the chapter and
highlights the areas for further work.

2. Knowledge-based continuous quality improvement in manufacturing

processes
Continuous quality improvement is an important concept in the field of manufacturing
quality management. DMAIC (Define-Measure-Analyze-Improve-Control) for six-sigma is
the most commonly used model of continuous quality improvement. Fig.1 illustrates the
processes of DMAIC.

Fig. 1. The processes of DMAIC

We can observe from Fig. 1 that DMAIC is a problem driven approach. The entire process
begins with locating a problem. However, in a complicated manufacturing process, such as
semiconductor manufacturing and steelmaking, it may not be easy to identify and define a
proper problem. Moreover, in high-speed manufacturing processes quality problems must
be quickly identified and eliminated. Otherwise it may lead to a large amount of loss in both
cost and productivity. Therefore, we propose a knowledge-based quality improvement
model (see Fig. 2). Different from DMAIC, this model is a goal-driven process. The central
idea of the knowledge-based quality improvement mode is to mine the mass data collected
from a manufacturing process using automated data mining techniques. The goal is to
improve the quality performance of manufacturing processes by quickly identifying and
eliminating quality problems.
In the knowledge-based quality improvement model, the first step is to define the goal. The
goal here may be defect elimination, efficiency improvement, or yield improvement. Data
mining is used to analyze the quality related data for finding the knowledge between the
goal and the factors such as machinery parameters, operators, and material vendors. After

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 359

the knowledge has been verified, opportunities of quality improvement can be identified
using the knowledge and patterns learned by data mining techniques. The scope of the
problem can be broad across different phases of a manufacturing process. In the following
sections, we explained how to apply the model to parameter optimization, quality diagnosis
and service data analysis.

Fig. 2. The knowledge-based quality improvement mode

3. Using data mining to improve quality in manufacturing processes

There are three important processes in manufacturing enterprises, including design,
manufacturing and service. Data mining techniques can be used in all three processes to
improve the quality of manufacturing processes and final products.

3.1 Parameter optimization with data mining

Parameter optimization is critical to quality improvement in manufacturing processes.
There are various methods proposed for parameter optimization. DOE (Design Of
Experiments) is considered as the most important one (Myers R H, 1985). RSM (Response
Surface Methodology) is a technique developed based on DOE and has been most widely
used. The main idea of DOE and RSM is to build a function between the inputs (factors) and
outputs (responses) of a process. The parameters of the function can be optimized using
mathematical methods.
Although DOE is valuable for parameter optimization, there are also drawbacks. DOE is a
static method in which the parameters are optimized in a certain situation and the
parameters cannot be adjusted during the real manufacturing process that is often dynamic
in all kinds of situations. In addition the DOE model only considers a small number of
factors with the constriction of cost.
To overcome the problems of DOE, we propose a technique that combines DOE and data
mining. DOE is used for parameter optimization for certain static modes. Data mining is
employed to analyze actual quality measures with different parameters settings. It aims to
learn the patterns between the parameters settings and quality outcome. The pattern can
then be used to dynamically adjust the parameters during the manufacturing process so that
undesired outcome can be avoided. In the following, we use an injection modeling process
in semiconductor assembly as an example to illustrate the process.
In the molding process of LED packaging, there can be many factors affecting the quality of
final products. We chose four most important factors with screening experiments. The
factors are explained below.
1. Mold temperature (°C). The mold temperature will affect the shape of final products,
which is a critical quality characteristic. If the temperature is too high, the number of
products with defects will increase.

www.intechopen.com
360 Data Mining and Knowledge Discovery in Real Life Applications

Table 1. The RSM table for parameter optimization

2. Warm-up temperature (°C). In injection molding process, epoxy compound, which is a
kind of material, must be softened before being used. The warming up temperature is
to ensure the material be softened in a proper temperature. The warm-up temperature
is also very important to the process.
3. Screw pressure (Newton). If the screw pressure is too high, the wire attached to the dice
and the PCB will be skewed. On the other hand, if the screw pressure is too low, the
epoxy compound will be cooled and also lead to quality problems.
4. Screw duration (Second.). Screw duration is the time intervals between the beginning
and the end of molding processes. The epoxy compound will be cooled with too long
screw duration. Screw duration is also a vital parameter of this process.
Experimental data are collected in a RSM table (Table 1) and analyzed using the Minitab©

The mold temperature is 142.3 °C, the warm-up temperature is 65.7 °C, the screw pressure is
software. And the optimized manufacturing parameters are found using statistical methods.

22.4 Newton and the screw duration is 40 seconds.

It must be pointed out that these optimized parameters are obtained in a controlled
experiment. The parameters may not achieve the optimal results in real manufacturing

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 361

conditions. Other uncontrollable factors may also have an effect on the quality of final
products. Thus the process should be monitored and optimized continuously using real
quality related data. Data mining can be used to serve that purpose by automatically
obtaining knowledge and patterns about the manufacturing process.
We collect 1000 records randomly from the molding processes. Each record consists of
values of the four factors. The records were classified based on their defect rates. Those
records whose defect rates were higher than 3.0% were categorized into a negative class and
labeled with L. All other records belonged to a positive class and were labeled with H. We
use the C5.0 decision tree algorithm to analyze the data. Fig. 3 shows the result of the
decision tree. We can observe that the right branch of the tree achieves a better performance
than the left branch. The path indicated by the circled nodes provides us guidance for
parameter optimization.
The decision tree shown in Fig. 3 can also be presented by a set of rules. We have identified
five rules. Two of them lead to the classification of the positive class (H) where the other
three predict the negative class (L).
1) Rule 1 for H (144; 0.973)
If Mold_Tem > 135.345 and
Warm_Tem > 66.762 and
Pressure > 21.235
Then H.
2) Rule 2 for H (127; 0.961)
If Mold_Tem <= 135.746 and
Warm_Tem > 62.106 and
Warm_Tem <= 66.762 and
Pressure > 20.733
Then H.
3) Rule 1 for L (44; 0.848)
If Mold_Tem <= 135.345 and
Warm_Tem > 66.761
Then L.
4) Rule 2 for L (390; 0.625)
If Pressure <= 21.235
Then L.
5) Rule 3 for L (790; 0.598)
If Warm_Tem <= 66.762
Then L.
Although a decision tree does not provide precisely optimized parameters like what the
RSM method does, decision tree can analyze a very large amount of quality related data
with noise which is still a constriction in DOE. The combination of these two methods can
provide more feasible results than using DOE only.

3.2 Quality diagnosis with data mining

During a manufacturing process, product quality can be affected by two types of variation:
random variations and assignable variations. Random variations are caused by the intrinsic
characteristics of a manufacturing process and cannot be eliminated completely. Assignable
variations are often produced by a flawed manufacturing setup that involves machinery,

www.intechopen.com
362 Data Mining and Knowledge Discovery in Real Life Applications

Fig. 3. The decision tree of manufacturing data analysis

operators, materials and environment. Assignable variations are predictable and such can be
eliminated as soon as we identify them.
SPC (Statistical Process Control) and MSPC (Multivariate SPC) are the most widely used
tools in manufacturing for finding assignable variations. Although they can effectively

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 363

detect assignable variations in manufacturing processes, they give no clue to identifying the
root causes of the assignable variations. Data mining techniques can again be employed in
this case to provide insights for quality diagnosis.

Fig. 4. The combination of SPC and data mining

In the model shown in Fig. 4, the yield ratio of a product is defined as the index of the
quality performance of a manufacturing process. The chart on the left is a control chart.
When the chat shows alarming signals, i.e., points located beyond the control limit, the data
mining process will be engaged. Data related to quality are stored in a data warehouse. Data
mining techniques such as a decision tree and association rule mining can be applied to the
data to identify the causes of the alarming signals.

Fig. 5. A mobile phone assembly line model

We use a mobile phone assembly line as an example to explain the process described above.
Electronic test is an important process in a mobile phone assembly lien. During electronic
test, all the products will be tested using an electronic test instrument. Quality related data

www.intechopen.com
364 Data Mining and Knowledge Discovery in Real Life Applications

can be collected off the test instrument and stored in a database. We can then analyze the
data using data mining in order to find ways to improve the quality of the manufacturing
process. A mobile phone assembly line model is presented in Fig. 5. There are m assembly
lines while products assembled in these lines are randomly distributed to n testers. The test
results are classified into two classes: Pass and Fail.

Fig. 6. The distribution of NBF values

In a mobile phone assembly line, there is a very important process named electronic test. In
this process, all the products will be tested with an electronic test instrument. And the data
are collected and stored in a database. We can analyze the data with data mining to find the
opportunity to improve the quality of manufacturing process. The mobile phone assembly
line is illustrated in fig. 5. There are m assembly lines and the products assembled in these
lines are randomly distributed to n testers. The test results are classified into two classes,
marked with Pass and Fail. Furthermore, there are also 27 numerical attributes in the test
results, but they are not concerned in this study.
Testers are assumed to work under the same condition with same precision. The number
between failures (NBF), i.e., the number of products that passed the test between two
successive failure tests, is used to measure the quality performance of the manufacturing
process. A larger NBF denotes a more stable manufacturing process.
In the experiment we collected 62,592 records from the electronic test process of the mobile
phone assembly line. The NBF values of the records are shown in Fig. 6 that provides a
concise description of the data.
We used Matlab software to analyze the distribution of NBF values. Fig. 7 shows that the
data can fit to a lognormal distribution. Fig. 7(a) shows the probability density function plot
while Fig. 7(b) draws the probability plot of the NBF values and the lognormal distribution
curve. The parameter estimation of the lognormal distribution is presented in Table 6.

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 365

Fig. 6. Distribution fitting of the NBF data

Table 6. The estimated parameters of the lognormal distribution

The standard error of both the mean and the standard deviation of the NBF data are really
small. Therefore we can say that the NBF data can approximately fit a lognormal
distribution with mean=3.54087 and sigma=0.70609. With the estimated parameters, we can
get the 0.0027 fractile, which is a vital parameter in SPC, of the lognormal distribution is
4.78. This means that if the NBF is smaller than 5, we can make the decision that there are
assignable variations in the process and the expected probability of an error decision is less

www.intechopen.com
366 Data Mining and Knowledge Discovery in Real Life Applications

than 0.27%. Thus we can set the CL=5 as a control line of the SPC chart for the NBF data. If
there are points distributed beyond the control limits, the data mining methods will be used
for quality diagnosis.
Next an association rule mining tool is used to find the root cause of the assignable
variations of the process. Attributes such as tester ID, assembly line number, and test results
are supplied into our association rule mining tool with test results being the consequent
variable. We used the Apriori algorithm as the association rules analysis method. The
minimum antecedent support was set to 0.65% while the minimum rule confidence was
80%. The obtained association rules are presented in Table 7.

Table 7. The obtained association rules

Some conclusions drawn from the association rules are listed as below.
1. There were 3,810 mobile phones assembled by line 15, 99.948% of which failed the test
(Rule 1).

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 367

2. There were 4,209 mobile phones assembled by line 20. 99.81% of which failed the test
(Rule 2). Rules 3- extended Rule 2 by considering each different tester.
3. There were 3,216 mobile phones assembled by line 33, all of which failed the test (Rules
10 and 11).
4. There were 4,297 mobile phones assembled by line 36, all of which failed the test (Rules
12-20).
5. There were 585 products tested by tester 104, 83.077% of which failed the test (Rule 21).
In all the 62,592 products, there were 16,905 products that failed the test. The percentage of
the failed tests was 27%. The failed products assembled by lines 15, 20, 33, 36 or tested by
tester 104 are summed up to 16,117. That means nearly 95% of the products that failed in the
tests were caused by the four assembly lines and/or tester 104. This example shows how
data mining techniques can be used to identify the root causes of quality problems in a
manufacturing process. That kind of knowledge is valuable for quality diagnosis and
quality improvement.

3.3 Quality improvement using customer service data

Manufacturing companies have obligations to repair sold products covered by the
manufacturer’s warranty. Customer service records can be a valuable information source for
quality improvement. Those records usually are multi-dimensional with many attributes. It
is a challenging issue for companies to extract useful information from the vast amount of
customer service records and provide feedback to product design and manufacturing
quality improvement.
We propose a research design for customer service data analysis. It includes four major
steps as we discuss below.
Step 1: Data structure design
Customer service data usually includes customer ID, product ID, service items and duration
of the owner ship. Fig. 7 shows an example of the relational database used to store the
customer service data. The database design follows the guidelines of developing a relational
database (Edgar F Codd, 1970).
This is the fact table of customer service data. We can define different dimensions for data
analysis. For example, in a mobile phones customer service domain, dimensions are
customer, product, service items and the duration of ownerships’.

Fig. 7. The database design for the customer service data

www.intechopen.com
368 Data Mining and Knowledge Discovery in Real Life Applications

Step 2: MDA (Multi-Dimensional Analysis)

The stored customer data can be analyzed in different dimensions. And the MDA can give a
statistical interpretation of the data. With the drill-down and drill-up operations, we can
analyze the data at a detailed level or a summarized level.
Step 3: Data mining
With the integrated customer service data, data mining techniques such as decision tree or
association rule mining can be used to find the relationship among customers, products,
durations and service items. These kinds of relationships will help improve product
design.
Step 4: Rules certification and usage
The patterns and knowledge obtained by data mining techniques must be verified by
domain experts or customer interview before being applied.
We illustrating the methods discussed above using 2000 records extracted from a mobile
phones manufacturer’s service process. The purpose is to demonstrate how the learned
knowledge can improve the quality of products by design.
The data were rearranged to a table with seven columns, including customer ID, Age,
District, Gender, Product Type, Service Item and Duration. The customer’s age was divided
into five intervals of above 55, between 45 and 55, between 35 and 45, between 25 and 35
and below 25. The location of customers was divided into four districts; they are southeast,
southwest, northeast and northwest of China. There are four types of products named as
series A, series B, series C and series D. The service items were classified into four types
including the product appearance, product function, product performance, and others. The
duration of ownerships was divided into four intervals: 1-6 months, 7-12 months, 13-18
months and 19-24 months.
To identify the relationship among these factors, we used the CHAID decision tree
algorithm (J. A. McCarty, 2007) to analyze the data. Fig. 8 shows the decision tree.
There is an interesting pattern in the decision tree. 48.3% of the products had function
defects. If the products were used in the district of southeast of China, the percentage
increased to 79.13%. Among them, 69.95% were the product of series C, 21.7% series B, and
only 8.3% product of series D. Obviously we can say that if the products of series C are used
in southeast of China, they are inclined to have functional defects.
Furthermore, we should analyze the rule deeply. After a technical examination of the
repaired products, it was found that the repaired products have almost the same erosion in
circuit boards. Compared with the environment conditions, it was found that the reason of
the functional defects of the products was the humid atmospheric corrosion. Erosion
occurred more easily on the products of series C than other types of products. The defects
were eliminated by design improvement. But this kind of problems was not very easy to
identify without the deep analysis of the service data.

4. System framework
Based on the above discussion, we propose a system infrastructure for quality improvement
using data mining techniques (Fig. 9). There are three layers in the model, namely data
collection layer, data analysis layer and data view layer. The function of each layer is
described below.

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 369

Fig. 8. The decision tree of service data analysis

1. Data collection layer
Functions in this layer are mainly used by operators for daily operations. They include
quality data maintenance, simple data analysis and query, and generating daily reports.
2. Data analysis layer
Functions of this layer are mainly used by mid-level managers and quality engineers. This
layer is the center of the system. Most of the quality data analysis functions are provided in
this layer including SPC, statistical analysis, data mining, quality diagnosis, etc.
3. Data view layer
Functions of this layer are mainly used by top managers for an integrated data view. In this
layer, data are often shown in graphs or reports.
Among the four layers, the data collection layer is the basis of the system. All the quality
data are collected and stored by this layer. The data analysis layer is the middle layer, where
the raw data are processed and analyzed. Results of the data analysis layer can be
transferred to the data view layer where can help high-level managers make quality
improvement decisions.

www.intechopen.com
370 Data Mining and Knowledge Discovery in Real Life Applications

Fig. 9. A system infrastructure for quality improvement using data mining techniques

5. Conclusions and further research

In a competitive global market, manufacturing enterprises must stay agile when making
quality improvement decisions. The development of IT and other related technologies
makes the collection of quality related data easy and cost-effective. However, it is still an
open question on how to leverage the large amount of quality data to improve
manufacturing quality. This chapter has approached the problem of quality improvement in
manufacturing processes using data mining techniques.
Firstly, we proposed a knowledge based six-sigma model where DMAIC for six-sigma was
used along with data mining techniques. The knowledge learned by the data mining
techniques was helpful in identifying potential quality problems and assisting quality
diagnosis. We also examined the problem of parameter optimization by applying data
mining techniques to DOE. A decision tree was build in order to dynamically adjust
parameter optimization. In addition, we applied data mining to quality diagnosis where an
association rule mining technique was used to analyze the electronic test data. The rules
obtained by data mining provided a direct guidance in identifying the root causes of the
quality problems. And the findings were beneficial for quality diagnosing that is still a
difficult problem in six-sigma. Furthermore, a decision tree was also used in the service data
analysis. The findings were valuable to improve product design. Finally, we presented a
system infrastructure for quality improvement in manufacturing processes.
On the other hand, there are still a lot of open questions to be studied in this field. Firstly,
the manufacturing processes are quite different with each other. The quality of the collected
data from each manufacturing process varies significantly. There can be a significant portion
of missing values and errors in the raw data. It is still a challenging issue as how to

www.intechopen.com
Quality Improvement using Data Mining in Manufacturing Processes 371

preprocess raw data for data mining algorithms. Secondly, patterns and knowledge learned
by data mining techniques are not always usable. How to ascertain the usable knowledge in
a large amount rules and patterns is also a problem that deserves attention. Finally, the
learned rules and patterns have to be analyzed by domain experts with their domain
knowledge. How to present the domain knowledge and build an automated knowledge
ascertain system is also a challenging issue in this field.

6. Acknowledgement
The authors of this work would like to thank the National Natural Science Foundation of
China (NSFC) who sponsored this research (grant no. 70572044). The authors would also
like to thank the anonymous reviewers for their constructive comments on this work.

7. References
Andrew Kusiak, Christian Kurasek. Data mining of printed-circuit board defects, IEEE
transactions on robotics and automation, vol.17, No.2, 2001.
Angie Patterson, Piero Bonissone, Marc Pavese. Six sigma applied throughout the lifecycle
of an automated decision system, quality and reliability engineering international,
2005.21: 275-292
Ben Khediri Issam, Limam Mohamed. Support vector regression based residual MCUSUM
control chart for autocorrelated process, Applied mathematics and computation,
2008 (In press)
Chen-Fu Chien, Wen-Chih Wang, Jen-Chieh Cheng. Data mining for yield enhancement in
semiconductor manufacturing and an empirical study, Expert Systems with
Applications, 2007.33: 192-198
Chen-Fu Chien, Huan-Chung Li and Angus Jeang. Data mining for improving the solder
bumping process in the semiconductor packaging industry, Intelligent systems in
accounting, finance and management, 2006.14: 43-57
Edgar F. Codd, A Relational Model of Data for Large Shared Data Banks, Communications
ofthe ACM 13(6): 377-387.
Giovanni C Porzio, and Giancarlo Ragozini. Visually mining off-line data for quality
improvement, Quality and reliability engineering international, 2003.19:273-283
Hsu-Hwa Chang. A data mining approach to dynamic multiple responses in Taguchi
experimental design, Expert systems with applications, 2008.35: 1095-1103
Ian H Witten and Eibe Frank, Data Mining: Practical Machine Learning Tools and
Techniques, Morgan Kaufmann Publishers, 2005.
J A McCarty, Manoj Hastak. Segmentation approaches in data-mining: a comparison of
RFM, CHAID, and logistic regression, Journal of Business Research, 2007.60: 656-
662.
Kaidi Zhao, Bing Liu, Tomas M Tirpak and Weimin Xiao. A visual data mining frame work
for convenient identification of useful knowledge, proceedings of the fifth IEEE
international conference on data mining, 2005.
Myers R H, Montgomery D C. Response surface methodology. John Wiley & Sons, New
York, 1985.
Mu-Chen Chen. Ranking discovered rules from data mining with multiple criteria by data
envelopment analysis, Expert systems with application, 2007.33:1110-1116

www.intechopen.com
372 Data Mining and Knowledge Discovery in Real Life Applications

Rakesh Menon, Loh Han Tong, S Sathiyakeerthi, Aarnout Brombacher and Christopher
Leong. The needs and benefits of applying textual data mining within the product
development process, Quality and reliability engineering international, 2004.21:1-15
Ruey-Shiang Guh, Yeou-Ren Shiue. An effective application of decision tree learning for on-
line detection of mean shifts in multivariate control charts, Computers & Industrial
Engineering, 2008 (In Press)
Seyed Taghi Akhavan Niaki, Babak Abbasi. Fault diagnosis in multivariate control charts
using artificial neural networks, Quality and reliability engineering international,
2005.21: 825-840
Sébastien Gebus, Kauko Leiviskä. Knowledge acquisition for decision support systems on
an electronic assembly line, Expert systems with applications, 2007 (In press).
Shao-Chuang Hsu, Chen-Fu Chien. Hybrid data mining approach for pattern extraction
from wafer bin map to improve yield in semiconductor manufacturing, Int J
Production Economics, 2007.107: 88-103
Tai-Yue Wang, Long-hui Chen. Mean shifts detection and classification in multivariate
process: a neural-fuzzy approach, Journal of intelligent manufacturing, 2002.12:
211-221

www.intechopen.com
 Data Mining and Knowledge Discovery in Real Life Applications
Edited by Julio Ponce and Adem Karahoca

ISBN 978-3-902613-53-0
Hard cover, 436 pages
Publisher I-Tech Education and Publishing
Published online 01, January, 2009
Published in print edition January, 2009

This book presents four different ways of theoretical and practical advances and applications of data mining in
different promising areas like Industrialist, Biological, and Social. Twenty six chapters cover different special
topics with proposed novel ideas. Each chapter gives an overview of the subjects and some of the chapters
have cases with offered data mining solutions. We hope that this book will be a useful aid in showing a right
way for the students, researchers and practitioners in their studies.

How to reference
In order to correctly reference this scholarly work, feel free to copy and paste the following:

Shu-guang He, Zhen He, G. Alan Wang and Li Li (2009). Quality Improvement using Data Mining in
Manufacturing Processes, Data Mining and Knowledge Discovery in Real Life Applications, Julio Ponce and
Adem Karahoca (Ed.), ISBN: 978-3-902613-53-0, InTech, Available from:
http://www.intechopen.com/books/data_mining_and_knowledge_discovery_in_real_life_applications/quality_im
provement_using_data_mining_in_manufacturing_processes

InTech Europe InTech China

University Campus STeP Ri Unit 405, Office Block, Hotel Equatorial Shanghai
Slavka Krautzeka 83/A No.65, Yan An Road (West), Shanghai, 200040, China
51000 Rijeka, Croatia
Phone: +385 (51) 770 447 Phone: +86-21-62489820
Fax: +385 (51) 686 166 Fax: +86-21-62489821
www.intechopen.com
© 2009 The Author(s). Licensee IntechOpen. This chapter is distributed
under the terms of the Creative Commons Attribution-NonCommercial-
ShareAlike-3.0 License, which permits use, distribution and reproduction for
non-commercial purposes, provided the original is properly cited and
derivative works building on this content are distributed under the same
license.