Economic and Financial Knowledge-Based Processing

by Louis F. Pau, Claudio Gianotti

Overview

As banks, financial services, insurances, and economic research units worldwide strive to add knowledge based capabilities to their analyses and services, or to create new ones, this volume aims to provide them with concrete tools, methods and application possibilities. The tutorial component of the book relies on case study illustrations, and on source code in some of the major artificial intelligence languages. The applications related component includes an extensive survey of real projects, and a number of thorough generic methods and tools for auditing, technical analysis, information screens and natural-language front-ends. The research related component highlights novel methods and software for economic reasoning under uncertainty and for fusion of qualitative/quantitative model-based economic reasoning.

Product Details

ISBN-13:	9783642760044
Publisher:	Springer Berlin Heidelberg
Publication date:	09/11/2013
Edition description:	Softcover reprint of the original 1st ed. 1990
Pages:	366
Product dimensions:	6.69(w) x 9.61(h) x 0.03(d)

Table of Contents

A reader’s guide.- 0 Introduction.- 0.1 Introduction.- 0.2 The strategic challenge to banks and insurance.- 0.3 The strategic challenge to financial services.- 0.4 The strategic challenge to economic analysis and decision making.- 0.5 The strategic challenge for business management.- 0.6 Conclusion.- 1 Basic concepts.- 1.1 Introduction.- 1.2 Survey of AI applications in finance and economics.- 1.3 Case studies and examples.- 1.4 The mortgage loan credit granting case study.- 1.4.1 Problem statement.- 1.4.2 Knowledge base.- 1.4.3 Unification.- 1.4.4 Probability.- 1.4.5 Inference.- 1.4.6 Explanations.- 1.4.7 Knowledge acquisition.- 1.4.8 Expert system architecture.- 1.4.9 Risk analysis inference control structure.- 1.5 AI and Decision support.- 2 Applications of Artificial Intelligence in banking, financial services and economics.- 2.1 The motivations for the use of AI.- 2.2 Survey of development projects.- 2.3 Development and delivery environments.- 2.4 Generic domain utilities.- 2.5 Inference control and conflict resolution strategies.- 2.6 Table of projects.- 2.7 Project references.- 3 Knowledge Representation.- 3.1 Introduction.- 3.1.1 Motivation.- 3.1.2 Explicit vs. implicit knowledge.- 3.1.3 The knowledge representation problem.- 3.1.4 Knowledge for economic/financial reasoning.- 3.1.5 Sources of economic/financial knowledge.- 3.1.6 Knowledge representation languages and formal languages.- 3.1.7 Segmentation of knowledge types for problem-solving.- 3.1.8 Fundamental knowledge representation formalisms.- 3.1.9 Classification criteria for knowledge representation languages.- 3.1.10 Adequacy of knowledge representation formalisms.- 3.2 Case study: a tax adviser.- 3.2.1 Structure of a tax form.- 3.2.2 Representation guidelines.- 3.2.3 Knowledge representation formalisms.- 3.3 The graph and tree data structures.- 3.3.1 Motivation.- 3.3.2 Graphs.- 3.3.3 Trees.- 3.4 Semantic networks.- 3.4.1 Motivation.- 3.4.2 Causality networks.- 3.4.3 Application: a simple economic model.- 3.4.4 Dependency graphs.- 3.5 Logic.- 3.5.1 Motivation.- 3.5.2 An introduction to predicate calculus.- 3.5.2.1 Logic connectives.- 3.5.2.2 Quantifiers.- 3.5.2.3 Model theory.- 3.5.3 Guidelines for logic-based knowledge representation.- 3.5.4 Logic inference.- 3.5.5 Clausal logic and resolution.- 3.5.6 Logic and semantic networks.- 3.5.7 Application: representing part of the Italian fiscal regulation.- 3.5.8 Pros and cons of logic.- 3.6 Rules.- 3.6.1 Motivation.- 3.6.2 Facts.- 3.6.3 Rules.- 3.6.4 Rules as a knowledge representation formalism.- 3.6.5 Applications of rule-based representation.- 3.6.6 Reasoning with rules.- 3.6.7 The inference engine.- 3.6.8 Metarules.- 3.6.9 Rules vs. procedural programming.- 3.6.10 Rules vs. logic.- 3.6.11 Concluding remarks.- 3.7 Frames.- 3.7.1 Motivation.- 3.7.2 Frames, slots and facets.- 3.7.3 Procedural attachment.- 3.7.4 Interpretations of frames.- 3.7.5 Taxonomies.- 3.7.6 Hierarchical networks.- 3.7.7 Other relations among frames.- 3.7.8 Comparative descriptions.- 3.7.9 Inheritance.- 3.7.10 Inheritance mechanisms.- 3.7.11 Frames and semantic networks.- 3.7.12 Frames vs. logic.- 3.8 Temporal reasoning.- 3.8.1 Introduction.- 3.8.2 Temporal logic.- 3.8.3 Time-interval reasoning.- 3.8.4 Temporal constraints.- 3.8.5 Feature extraction in the time domain.- 3.8.6 Temporal inference.- 4 Artificial Intelligence Programming Languages.- 4.1 Introduction.- 4.1.1 Syntax and semantics.- 4.1.2 AI programming languages.- 4.1.3 Symbols and symbolic expressions.- 4.1.4 Interactivity and language interpreters.- 4.1.5 A classification of programming languages.- 4.2 Language syntax and parsing.- 4.2.1 Language syntax.- 4.2.2 Language parsing.- 4.2.3 Context-free and context-sensitive languages.- 4.3 LISP.- 4.3.1 A first look.- 4.3.2 Atoms and lists.- 4.3.3 Evaluation rules for atomic expressions.- 4.3.4 LISP functions.- 4.3.5 Functional composition and abstraction.- 4.3.6 Application: computing elasticities in economics.- 4.3.7 Functional vs. procedural programming.- 4.3.8 Boolean functions, IF and COND.- 4.3.9 Symbolic data structures.- 4.3.10 Assignment and evaluation of the data.- 4.3.11 Properties and association lists.- 4.3.12 Dynamic data typing.- 4.3.13 Identity of programs and data.- 4.3.14 Application: computing compound interests by recursion.- 4.4 Prolog.- 4.4.1 Beliefs in Prolog.- 4.4.2 Facts.- 4.4.3 Rules.- 4.4.4 Goals.- 4.4.5 Structured objects: tuples, lists and trees.- 4.4.6 Parse trees of Prolog expressions.- 4.4.7 Pattern-matching and unification.- 4.4.8 Infinite trees.- 4.4.9 Recursion.- 4.4.10 The inference engine.- 4.4.11 Controlling backtracking:!.- 4.4.12 Identity of data and programs.- 4.4.13 Application: a Prolog knowledge-based tax adviser.- 4.5 Object-oriented programming.- 4.5.1 Introduction.- 4.5.2 Object-oriented programming concepts.- 5 Search and causal analysis.- 5.1 Motivation.- 5.2 State-based representation of problems.- 5.3 Problem graphs.- 5.3.1 Implicit representation of graphs and trees.- 5.4 Search and knowledge.- 5.5 Search procedures.- 5.5.1 A generic search procedure.- 5.5.2 Classification of search methods.- 5.5.3 Application-specific search and mixed procedures.- 5.5.4 Optimization criteria.- 5.6 Application: a simple economic model in graph form.- 5.7 Simple propagation.- 5.8 Propagation with alternatives: depth-first.- 5.8.1 Representation of directed graphs.- 5.8.2 A description of the depth-first algorithm.- 5.8.3 An implementation of depth-first algorithm.- 5.8.4 Examples.- 5.9 Introducing side effects: breadth-first.- 5.9.1 A description of the breadth-first algorithm.- 5.9.2 An implementation of breadth-first algorithm.- 5.9.3 Examples.- 5.10 Case study: causal analysis in linear economic models.- 5.10.1 Causality representation in economic models.- 5.10.2 Causal analysis of a simple economic model.- 5.10.3 Causal ordering.- 5.10.4 Algorithm for causal assignment.- 5.10.5 An algorithm for causal analysis and consistency.- 5.11 Heuristic search methods.- 5.11.1 Hill-climbing algorithm.- 5.11.2 Beam-search algorithm.- 5.11.3 Best-first algorithm.- 5.11.4 A* algorithm.- 6 Neural processing and inductive learnings.- 6.1 Introduction.- 6.2 Neural processing for learning and classification.- 6.2.1 Neural models.- 6.2.2 Neural learning.- 6.2.3 Neural learning algorithms.- 6.2.4 Consultation.- 6.2.5 Performance evaluation.- 6.3 Inductive learning.- 6.3.1 Introduction.- 6.3.2 Concept learning.- 6.3.3 Induction algorithms.- 6.3.4 ID3 induction of decision trees.- 6.3.5 Examples.- 6.4 Extensions to neural processing.- 6.4.1 Neural decision logic.- 6.4.2 Learning how to forecast.- 6.4.3 Other applications.- 7 Technical analysis for securities trading.- 7.1 Introduction.- 7.2 Curve generation by a syntactic grammar.- 7.3 Curve segmentation.- 7.4 Segmentation of noisy curves.- 7.5 Analysis evaluation rules.- 7.6 Technical analysis on several curves and software implementation.- 7.7 Time series analysis.- 7.8 Examples of concurrent trading rules.- 7.9 Off-line analysis for learning.- 7.10 Forecasting.- 7.11 Trade generation.- 8 Intelligent information screens.- 8.1 Introduction.- 8.2 Selective object-oriented data acquisition.- 8.3 Knowledge-based information screens.- 8.4 Knowledge-based filters for financial information screens.- 8.5 Information retrieval aspects.- 8.6 Data fusion.- 8.7 Correlation.- 9 Natural language front-ends to economic models.- 9.1 Introduction.- 9.2 Prolog parser for NL front-ends.- 9.3 Definite clause grammar in Prolog.- 9.4 Translation of DCG grammar rules into Prolog clauses.- 9.5 DCG parser.- 9.6 Reasoning from NL analysis.- 9.6.1 Form input.- 9.6.2 Validation of input.- 9.6.3 Modeling from NL analysis.- 9.6.4 Generation of temporal reasoning from NL analysis.- 10 Trade selection with uncertain reasoning on technical indicators.- 10.1 Introduction.- 10.2 The theory of Dempster-Shafer.- 10.2.1 Basic probability assignment.- 10.2.2 Gedibility belief and plausibility.- 10.3 Pooling evidence.- 10.4 Application: pooling evidence about trading.- 11 Currency risk management.- 11.1 Introduction: risk planning over time.- 11.2 Single period model.- 11.3 Multi-period model.- 11.4 Knowledge-based risk management.- 11.5 Risk allocation procedure.- 12 Reasoning procedures in knowledge-based systems for economics and management.- 12.1 Introduction.- 12.2 Objects in decision analysis.- 12.3 Classification of decision methods.- 12.3.1 Perception criterion.- 12.3.2 Rationality criterion.- 12.3.3 Action criterion.- 12.4 Logics and constraints.- 12.5 Truth maintenance as rational decision-making.- 12.6 Search over time and disequilibrium.- 12.7 Conflict resolution.- 12.8 Search over AND/OR graphs.- 12.9 Power relations and gaming for the selection of solutions.- 12.9.1 Game theory and AI.- 12.9.2 Definitions.- 12.9.3 Relations to search strategies.- Appendix 1 Software Codes.- Appendix 2 Predefined LISP and Prolog expressions 333.