Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Data Mining Practical Machine Learning Tools and Techniques

RFM stands for Recency, Frequency and Monetary value. RFM analysis is a marketing technique used for analyzing customer behavior such as how recently a customer has purchased (recency), how often the customer purchases (frequency), and how much the customer spends (monetary). It is a useful method to improve customer segmentation by dividing customers into various groups for future personalization services and to identify customers who are more likely to respond to promotions. In recent years, data mining applications based on RFM concepts have also been proposed for different areas such as for the computer security (Kim et al., 2010), for automobile industry (Chan, 2008) and for the electronics industry (Chiu et al., 2009). Research cases of data mining with RFM variables include different data mining techniques such as neural network and decision tree (Olson et al., 2009), rough set theory (Cheng & Chen, 2009), self organizing map (Li et al., 2008), CHAID (McCarty and Hastak, 2007), genetic algorithm (Chan, 2008) and sequential pattern mining (Chen et al., 2009; Liu et al., 2009). Integration of RFM analysis and data mining techniques provides useful information for current and new customers. Clustering based on RFM attributes provides more behavioral knowledge of customers’ actual marketing levels than other cluster analyses. Classification rules discovered from customer demographic variables and RFM variables provides useful knowledge for managers to predict future customer behavior such as how recently the customer will probably purchase, how often the customer will purchase, and what will the value of his/her purchases. Association rule mining based on RFM measures analyzes the relationships of product properties and customers’ contributions / loyalties to provide a better recommendation to satisfy customers’ needs. This chapter presents incorporating RFM analysis into data mining techniques to provide market intelligence. It proposes a new three-step approach which uses RFM analysis in data mining tasks, including clustering, classification and association rule mining, to provide market intelligence and to assist market managers in developing better marketing strategies. In our model, (i) once clustering task is used to find customer segments with similar RFM values, (ii) then, using customer segments and customer demographic variables, classification rules are discovered to predict future customer behaviors, (iii) finally; association rule mining is carried out for product recommendation. The proposed model depends on the sentence "the best predictor of future customer behavior is past customer behavior". (Swearingen, 2009) The purpose of this study is to provide better product recommendations than simple recommendations, by considering several parameters together: customer’s segment, the

/pdf/data-mining-using-rfm-analysis-3gh64ejxqz.pdf

Data Mining Using RFM Analysis

The introduction of computing and communications technologies within cars raises a range of novel human-computer interaction (HCI) issues. In particular, it is critical to understand how user-interfaces within cars can best be designed to account for the severe physical, perceptual and cognitive constraints placed on users by the driving context. This chapter introduces the driving situation and explains the range of computing systems being introduced within cars and their associated user-interfaces. The overall human-focused factors that designers must consider for this technology are raised. Furthermore, the range of methods (e.g., use of simulators, instrumented vehicles) available to designers of in-car user-interfaces are compared and contrasted. Specific guidance for one key system, vehicle navigation, is provided in a case study discussion. To conclude, overall trends in the development of in-car userinterfaces are discussed and the research challenges are raised.

Designing and Evaluating In-Car User-Interfaces

Recently, Content Delivery Networks (CDN) have become more and more popular. The technology itself is ahead of academic research in this area. Several dimensions of the technology have not been adequately investigated by academia. These dimensions include outline management, security, and standardization. Discovering and highlighting aspects of this technology that may have or have not been covered by academic research is the first step toward helping academia bridge the gap with industry or even go one step further to lead industry in the right direction. This suggests a comprehensive survey on research works in this regard. The literature in this area has already come up with some surveys and taxonomies, but some of them are outdated or do not cover every aspect of CDN while others fail to detect existing trends or to develop a holistic roadmap for research on the technology. Furthermore, none of the existing surveys aim at enlightening the dark aspects of the technology that have not been subject to academic research. In this survey, we first extract the lifecycle of a CDN as suggested by the existing research. Then, we investigate previous relevant works on each phase of the lifecycle to clarify where the research is currently located and headed. We show how CDN technology tends to converge with emerging paradigms such as cloud computing, edge computing, and machine learning, which are more mature in terms of academic research. This helps us determine the right direction for further research by revealing the deficiencies in existing works.

Content Delivery Networks: State of the Art, Trends, and Future Roadmap

In Thailand e-banking has been offered by various financial institutes including Thai commercial banks and government banks. However, e-banking in Thailand is not widely used and accepted as in other countries. Accordingly, the study of e-banking is scantly due to the limitation of data confidentiality. This study uses data mining techniques to analyse historical data of e-banking usages from a commercial bank in Thailand. These techniques including SOMS, K-Mean algorithm and marketing techniques-RFM analysis are used to segment customers into groups according to their personal profiles and e-banking usages. Then Apriori algorithm is applied to detect the relationships within features of e-banking services. Typically, results of this study are presented and can be used to generate new service packages which are customised to each segment of e-banking users.

/pdf/clustering-e-banking-customer-using-data-mining-and-4lcp0w74bh.pdf

Clustering e-Banking Customer using Data Mining and Marketing Segmentation

Cloud services are based upon virtualization technology. When a user requests for resources for computing, the cloud service provider creates a virtual machine and allocate necessary resources such as CPU and RAM sufficient for usages to a user. The user does not need to be aware of the physical machines. When resource requirement increases due to the number of accesses and workload resulting in more allocated resources needed, the cloud service provider must ensure its quality of service (QoS) and the guaranteed response time to the provided services in order to maintain the customer satisfaction level and its reputation. To minimize the monitoring process overhead of the cloud service provider's data center administrators, an efficient auto-scaling mechanism is needed for efficiently managing the data center. A horizontal auto-scaling with process migration mechanism for cloud services with skewness algorithm is then proposed and implemented in a Java-based cloud simulation and tested in two different scenarios: threshold-based auto-scaling without skewness algorithm and process migration and one with skewness algorithm and process migration. The results show that the skewness algorithm and process migration can help smooth out the fluctuation of the number of virtual machines being spawned and then very soon be deleted resulting in the reduction of the overhead of such process.

http://ieeexplore.ieee.org/iel7/7731988/7748833/07748936.pdf

Horizontal auto-scaling and process migration mechanism for cloud services with skewness algorithm

In the past few years, the mobile phone’s performance has increased rapidly. According to IDC’s Worldwide Mobile Phone 2004-2008 Forecast and Analysis, sales of 2.5G mobile phones will drive market growth for the next several years, with sales of 3G mobile phones finally surpassing the 100 million annual unit mark in 2007. Future mobile phones can support more than 20,000 colors. With the advancements in functionality and performance of mobile phones, users will use them for all sorts of activities, and that will increase mobile content service requests. Currently, the pricing of mobile content service is up to each provider; typically they implement a fixed price called a market price because the providers do not have a formula to estimate the price according to the actual cost of their services. This article proposes a dynamic pricing model based on net cost for mobile content services.

Dynamic Pricing Based on Net Cost for Mobile Content Services

In the past, the Internet topology was produced by using connection data from Internet Service Provider (ISP). However, if ISPs didn't update connection data or didn't want publish this data, the Internet topology couldn't be produced. Furthermore, there was no relationships information available between the Autonomous System (AS). In this paper, we use a technique, which does not need to use connection data from ISPs, to examine AS path in BGP routing table to produce Internet topology and infer the type of relationships between the ASs within Thailand.

A study of autonomous system relationships within Thailand

The advancement of technology to support data storage is easy to store with large volumes of data. In order to make data storage to be used extensively, the standard of data storage is formed. The World Health Organization defines numbers of the standard medical procedures to cover the all treatments without classifying any medical procedures by diseases. The selection of medical procedures is based on a patient's symptoms. Therefore, if the sets of medical procedures can identified, we may know the diseases of the patient or it can be used in disease surveillance. In addition, diabetes and hypertension are silent killers that have been threatening numbers of Thai people and also lead to many serious diseases. This research identified sets of medical procedures related to diabetes and/or hypertension using C4.5 and Naive Bayes algorithms. The results showed that C4.5 could identify sets of medical procedures related to Diabetes and/or Hypertension more effectively than the Naive Bayes algorithm.

Sukumal Kitisin

Papers

Clustering e-Banking Customer using Data Mining and Marketing Segmentation

Horizontal auto-scaling and process migration mechanism for cloud services with skewness algorithm

Dynamic Pricing Based on Net Cost for Mobile Content Services

A study of autonomous system relationships within Thailand

The classification of sets of medical procedures used in the treatment of Diabetes and/or Hypertension