Intelligent systems for money laundering prevention
Editor: Timon Molloy
Money Laundering Bulletin
Increasingly, Intelligent System software techniques are being incorporated into money laundering prevention systems. Such techniques allow automated systems to perform analysis and consider behaviours in more human and flexible ways. Intelligent systems apply a broad range of methods developed by the artificial intelligence and computer science research communities. These techniques have been applied to identify credit card fraud, insurance claims fraud, insider dealing, and money laundering and are also used in areas such as risk management, audit, marketing and customer relationship management (CRM). In the final part of a series of three articles Tony Wicks of Searchspace introduces important properties of intelligent systems and gives an overview of a number of techniques. The article then describes how such techniques can be best applied for money laundering prevention.
Adaptive Fuzzy Control: Experiments and Comparative Analyses
Authors: Ra´ul Ord Ünez, Jon Zumberge, Jeffrey T. Spooner, and Kevin M. Passino, Senior Member, IEEE
IEEE TRANSACTIONS ON FUZZY SYSTEMS, VOL. 5, NO. 2, MAY 1997
Index Terms Adaptive fuzzy control, fuzzy control, intelligent control.
Advances in nonlinear control theory have provided the mathematical foundations necessary to establish conditions for stability of several types of adaptive fuzzy controllers. However, very few, if any, of these techniques have been compared to conventional adaptive or nonadaptive nonlinear controllers or tested beyond simulation; therefore, many of them remain as purely theoretical developments whose practical value is difficult to ascertain. In this paper we will develop three case studies where we perform a comparative analysis between the adaptive fuzzy techniques in   and some conventional adaptive and nonadaptive nonlinear control techniques. In each case, the analysis will be performed both in simulation and in implementation, in order to show practical examples of how the performance of these controllers compares to conventional controllers in real systems.
Estimating Software Project Effort by Analogy Based on Linguistic Values
Authors: Ali Idri, Alain Abran, Taghi M. Khoshgoftaar
Estimation models in software engineering are used to predict some important attributes of future entities such as software development effort, software reliability and programmers productivity. Among these models, those estimating software effort have motivated considerable research in recent years. The prediction procedure used by these software -effort models can be based on a mathematical function or other techniques such as analogy based reasoning, neural networks, regression trees, and rule induction models. Estimation by analogy is one of the most attractive techniques in the s software effort estimation field. However, the procedure used in estimation by analogy is not yet able to handle correctly linguistic values (categorical data) such as 'very low', 'low' and 'high'. In this paper, we propose a new approach based on reasoning by analogy, fuzzy logic and linguistic quantifiers to estimate software project effort when it is described either by numerical or linguistic values; this approach is referred to as Fuzzy Analogy. This paper also presents an empirical validation of our approach based on the COCOMO'81 dataset.
Prototypes, Complements, and Fuzzy Work Practices: Assigning Causal Credit for Performance
Authors: Bruce Kogut, John Paul MacDuffie, and Charles Ragin
26 February 2002 First draft: April 1999
Reginald H. Jones Center working paper # 1999-08
The globalization of markets is marked by the faddish diffusion of complementary managerial practices that purport to lead to better performance. This complexity poses the problem of credit assignment: identifying the path of causality between interacting practices and their effects. We apply Ragin s (2000) fuzzy logic methodology to identify high performance configurations in the 1989 data set of MacDuffie (1995). The results indicate that discrete bundles of practices are associated with a prototypical understanding of best practices. Managerial discourse around prototypical best practices is the groping for better performance when the categorization and causality between practice and outcome are fuzzy.
A Web-based CBR Agent for Financial Forecasting
Authors: James N. K. Liu and Tommy T.S. Leung
Department of Computing, Hong Kong Polytechnic University, Hong Kong
Keywords: Web-based system, CBR agent, financial forecasting
This paper presents a Web-based case-based reasoning model to assist investors to determine stock trend signals for investment in stock business. The model conforms to a Webbased agent framework forming part of an advisory system for financial forecast. Much of the discussion will be devoted to the design and development of the framework and associated intelligent techniques. Different cases are collected based on the theory of waves features and their combinations. The agent framework supports processes including Knowledge generation, Wave units mining and Wave Pattern recognition, and Case Revise and Learning. Preliminary result indicates that the CBR agent model is promising and reasonable. It is feasible to capture the trading behavior of the market with expandable case options.
Intelligent hybrid systems for financial decision making
Author: Suran Goonatilake
Economics, Finance and Business
This case study describes the use of genetic-fuzzy hybrid systems for supporting financial decision making. A novel architecture for inducing fuzzy rule-bases using genetic algorithms is presented. This combination of genetic algorithms and fuzzy logic produces easy to understand transparent decision models that can be easily understood by technical personnel and high-level strategic decision makers alike. Although we discuss this approach with an example from the area of decision support in financial trading, this method evidently has wide applications in other areas of financial decision making including credit evaluation, corporate risk assessment, and insurance underwriting.
An Agent-Based Hybrid Intelligent System for Financial Investment Planning
Authors: Zili Zhang and Chengqi Zhang
School of Computing and Mathematics Deakin University, Geelong Victoria 3217, Australia
Faculty of Information Technology University of Technology, Sydney PO Box 123 Broadway, NSW 2007 Australia
Keywords: Multi-Agent Systems, Hybrid Intelligent Systems, Intelligent Agents, Distributed AI, Decision Making
Many complex problems such as financial investment planning involve many different components or sub-tasks, each of which requires different types of processing. To solve them, a great diversity of intelligent techniques including expert systems, fuzzy logic, neural networks, and genetic algorithms are required. These techniques are complementary rather than competitive and thus must be used in combination and not exclusively. This results in systems called hybrid intelligent systems. That is, hybrid solutions are crucial for complex problem solving and decision making. However, the design and development of hybrid intelligent systems is cult because they have a large number of parts or components that have many interactions. This paper describes an agent-based hybrid intelligent system for financial investment planning, which currently consists of 13 different agents. The experimental results show that all agents in the system can work cooperatively to provide reasonable investment advice. The success of the system indicates that agent technologies can significantly facilitate the building of hybrid intelligent systems, especially loosely coupled hybrid intelligent systems. The hybrid intelligent systems constructing from agent perspectives are very flexible and robust.
Statistical Fraud Detection: A Review
Authors: Richard J. Bolton and David J. Hand
Keywords: Fraud detection, fraud prevention, statistics, machine learning, money laundering, computer intrusion, e-commerce, credit cards, telecommunications.
Fraud is increasing dramatically with the expansion of modern technology and the global superhighways of communication, resulting in the loss of billions of dollars worldwide each year. Although prevention technologies are the best way of reducing fraud, fraudsters are adaptive and, given time, will usually find ways to circumvent such measures. Methodologies for the detection of fraud are essential if we are to catch fraudsters once fraud prevention has failed. Statistics and machine learning provide effective technologies for fraud detection and have been applied successfully to detect activities such as money laundering, e-commerce credit card fraud, telecommunication fraud, and computer intrusion, to name but a few. We describe the tools available for statistical fraud detection and the areas in which fraud detection technologies are most used.
The Theory of Fuzz Logic and its Application to Real Estate Valuation
Authors: Carlo Bagnoli, Halbert C. Smith
JOURNAL OF REAL ESTATE RESEARCH
VOLUME 16, NUMBER 2, 1998
Fuzzy logic is based on the central idea that in fuzzy sets each element in the set can assume a value from 0 to 1, not just 0 or 1, as in classic set theory. Thus, qualitative characteristics and numerically scaled measures can exhibit gradations in the extent to which they belong to the relevant sets for evaluation. This degree of membership of each element is a measure of the element s belonging to the set, and thus of the precision with which it explains the phenomenon being evaluated. Fuzzy sets can be combined to produce meaningful conclusions, and inferences can be made, given a specified fuzzy input function. The article demonstrates the application of fuzzy logic to an income-producing property, with a resulting fuzzy set output.
Data Mining Using Synergies Between Self-Organizing Maps and Inductive Learning of Fuzzy Rules
Authors: MARIO DROBICS, ULRICH BODENHOFER, WERNER WINIWARTER and ERICH PETER KLEMENT
Software Competence Center Hagenberg A-4232 Hagenberg, Austria
Identifying structures in large data sets raises a number of problems. On the one hand, many methods cannot be applied to larger data sets, while, on the other hand, the results are often hard to interpret. We address these problems by a novel three-stage approach. First, we compute a small representation of the input data using a self-organizing map. This reduces the amount of data and allows us to create two-dimensional plots of the data. Then we use this preprocessed information to identify clusters of similarity. Finally, inductive learning methods are applied to generate sets of fuzzy descriptions of these clusters. This approach is applied to three case studies, including image data and real-world data sets. The results illustrate the generality and intuitiveness of the proposed method.
Knowledge Modeling - State of the Art
Author: Vladan Devedzic
Department of Information Systems, FON - School of Business Administration University of Belgrade POB 52, Jove Ilica 154, 11000 Belgrade, Yugoslavia
A major characteristic of developments in the broad field of Artificial Intelligence (AI) during the 1990s has been an increasing integration of AI with other disciplines. A number of other computer science fields and technologies have been used in developing intelligent systems, starting from traditional information systems and databases, to modern distributed systems and the Internet. This paper surveys knowledge modeling techniques that have received most attention in recent years among developers of intelligent systems, AI practitioners and researchers. The techniques are described from two perspectives, theoretical and practical. Hence the first part of the paper presents major theoretical and architectural concepts, design approaches, and research issues. The second part discusses several practical systems, applications, and ongoing projects that use and implement the techniques described in the first part. Finally, the paper briefly covers some of the most recent results in the fields of intelligent manufacturing systems, intelligent tutoring systems, and ontologies.
On the Assessment of CommercialWebsite - An Expert System Approach
Authors: Xue Li and Wayne Huang
Proceedings of the Twelfth Australasian Conference on Information Systems
Keywords Expert Systems HA04, Software Evaluation ED02, Interactive Information Systems GE04
The assessment of commercial website is about the evaluation of websites on the Internet for commercial usage. From an information system perspective, they can be assessed objectively or subjectively. From a business point of view, they can be assessed quantitatively or qualitatively. In this paper we propose an expert system approach to evaluate commercial website in both of these two aspects in order to gain a clear picture of their values.
Fuzzy inductive reasoning, expectation formation and the behavior of security prices
Authors: Nicholas S.P. Tay, Scott C. Linn
Journal of Economic Dynamics & Control 25 (2001) 321}361
Keywords: Expectations; Learning; Fuzzy logic; Induction; Stock price dynamics
This paper extends the Santa Fe Arti"cial Stock Market Model (SFASM) studied by LeBaron, Arthur and Palmer (1999, Journal of Economic Dynamics and Control 23, 1487}1516) in two important directions. First, some might question whether it is reasonable to assume that traders are capable of handling a large number of rules, each with numerous conditions, as is assumed in the SFASM. We demonstrate that similar results can be obtained even after severely limiting the reasoning process. We show this by allowing agents the ability to compress information into a few fuzzy notions which they can in turn process and analyze with fuzzy logic. Second, LeBaron et al. have reported that the kurtosis of their simulated stock returns is too small as compared to real data. We demonstrate that with a minor modifcation to how traders go about deciding which of their prediction rules to rely on when making demand decisions, the model can in fact produce return kurtosis that is comparable to that of actual returns data. ( 2001 Elsevier Science B.V. All rights reserved.
A Hybrid Time Lagged Network for Predicting Stock Prices
Authors: S.C. Hui, M.T. Yap and P. Prakash
School of Applied Science, Nanyang Technological University Singapore
Keywords : Neural networks, hybrid approach, multilayer perceptron, Kohonen networks, stock prediction.
Traditionally, technical analysis approach, that predicts stock prices based on historical prices and volume, basic concepts of trends, price patterns and oscillators, is commonly used by stock investors to aid investment decisions. Advanced intelligent techniques, ranging from pure mathematical models and expert systems to neural networks, have also been used in many financial trading systems for predicting stock prices. In this paper, we propose the Hybrid Time Lagged Network (HTLN) which integrates the supervised Multilayer Perceptron using temporal back-propagation algorithm with the unsupervised Kohonen network for predicting the chaotic stock series. This attempts to combine the strengths of both supervised and unsupervised networks to perform more precise prediction. The proposed network has been tested with stock data obtained from the main board of Kuala Lumpur Stock Exchange (KLSE). In this paper, the design, implementation and performance of the proposed neural network are described.
Genetic Algorithms in Economics and Finance: Forecasting Stock Market Prices and Foreign Exchange ¾ A Review
Authors: Adrian E. Drake, Robert E. Marks
University of Stuttgart
AGSM University of New South Wales Sydney 2052 Australia
The increased availability of computing power in the past two decades has been used to develop new techniques of forecasting. Today's computational capacity and the widespread availability of computers have enabled development of a new generation of intelligent computing techniques, such as expert systems, fuzzy logic, neural networks, and genetic algorithms. These "intelligent" computing techniques are concerned with performing actions that imitate human decision makers. They are flexible and so can adapt to new circumstances. They can learn from past experience. Moreover, they can find solutions to problems unsolvable by traditional means. Of the computing techniques mentioned above, genetic algorithms are the most powerful and yet the most simple innovation. They are showing very promising results in improving or replacing existing statistical methods.
Consistent and complete data and expert mining in medicine
Authors: Boris Kovalerchuk, Evgenii Vityaev. James F. Ruiz
1Department of Computer Science, Central Washington University, Ellensburg, WA, 98926-7520, USA
2Institute of Mathematics, Russian Academy of Sciences, Novosibirsk 630090, Russia
3Department of Radiology, Woman s Hospital, Baton Rouge, LA 70895- 9009, USA
The ultimate purpose of many medical data mining systems is to create formalized knowledge for a computer-aided diagnostic system, which can in turn, provide a second diagnostic opinion. Such systems should be consistent and complete as much as possible. The system is consistent if it is free of contradictions (between rules in a computer-aided diagnostic system, rules used by an experienced medical expert and a database of pathologically confirmed cases). The system is complete if it is able to cover (classify) all (or largest possible number of) combinations of the used attributes. A method for discovering a consistent and complete set of diagnostic rules is presented in this chapter. Advantages of the method are shown for development of a breast cancer computer-aided diagnostic system
Building the Santa Fe Artificial Stock Market
Author: Blake LeBaron
Brandeis University June 2002
This short summary presents an insider s look at the construction of the Santa Fe artificial stock market. The perspective considers the many design questions that went into building the market from the perspective of a decade of experience with agent-based financial markets. The market is assessed based on its overall strengths and weaknesses.
Session 8PD Fuzzy Logic
Dr. Fred A. Watkins:
Las Vegas Spring Meeting May 22 24, 2000
Summary: The panelists cover: 1) The basic concepts underlying fuzzy logic 2) Current applications to risk classification and pricing 3) Future possible uses of fuzzy logic
Information Mining with the IBM Intelligent Miner Family
Author: Daniel S. Tkach
An IBM Software Solutions White Paper
Information mining refers to the process of extracting previously unknown, comprehensible, and actionable information from any source - including transactions, documents, e-mail, web pages, and other, and using it to make crucial business decisions. The two most pervasive types of information are structured data and text, therefore information mining includes data mining and text mining.The IBM Intelligent Miner Family that comprises the IBM Intelligent Miner for Data and the IBM Intelligent Miner for Text, provide the most advanced and comprehensive set of solutions for information mining in the industry. This document describes the information mining operations and techniques as they are implemented in the IBM Intelligent Miner Family, and highlights the applications that have proven to provide competitive advantages in many enterprises world-wide.
Soft Computing and Financial Engineering
Author: Arnold F. Shapiro
Smeal College of Business, Penn State University,
University Park, PA 16802, USA
Keywords: soft computing, financial engineering, neural networks, fuzzy logic, genetic, algorithms
One of the interesting subjects actuaries are (or should be) involved with is personal financial planning (PFP). On the liability side of PFP are the contingencies actuaries traditionally deal with and on the asset side are the financial risks associated with financial engineering. This study focuses on the soft computing aspects of the financial engineering portion of PFP. Soft computing (SC) has been applied to a number of systems in financial engineering, and in many cases has demonstrated better performance than competing approaches. The purpose of this research is to investigate the extent of these applications and to document the manner in which the SC technologies were implemented.