Main Page

NATIONAL UNIVERSITY OF SAN MARCOS

(University of Peru, DEAN OF AMERICA)

FACULTY OF ENGINEERING AND COMPUTER SYSTEMS

Academic Professional School of Systems Engineering

SYLLABUS

1. GENERAL SPECIFICATIONS

Course Name               : ARTIFICIAL INTELLIGENCE

Course Code                : 207008

Course Length              : 17 weeks

Form of Dictation         : Technical - experimental

Hours per week            : Theory: 3h – Labarotory: 2h

Nature                         : Vocational Training

Number of credits        : Cuatro (04)

Prerequisites                : 205007 –  Operations Research I

Academic Semester      : 2014-II

2. INTRODUCTION

Artificial intelligence is an area of ​​computer science that aims to develop the practical and theoretical foundations for the development of computer systems that have smart features and usually correspond to intractable problems. Many problems occurring in industry, services and entertainment are smart problems and its solution becomes increasingly essential in organizations due to the demands of competitiveness are increasing every day. 

Among the various types of intelligent systems are the human-machine games, knowledge-based systems, systems optimization and machine learning. 

In this course, an introduction to artificial intelligence and its applications in industry, services and entertainment is made, and shows how the artificial intelligence based technologies can create value and make organizations more competitive. Applications will also be displayed in software engineering. 

Discussed in more depth the design and implementation of human-machine gambling and knowledge-based systems.

3. COURSE ACHIEVEMENT

After completing the course the student will acquire a general knowledge of the area of ​​artificial intelligence, game design and implement human-machine competition based on artificial intelligence and knowledge-based systems, using clear and accurate search techniques in a space status and the CommonKADS methodology.

4. TEACHING UNITS

UNIT 1: Fundamentals of Artificial Intelligence

Achievement  The student at the end of the unit is understood that the artificial intelligence (AI) and its difference with information systems, some applications in industry and services, and the difficulty to solve through complexity theory problems.  Agenda  Classification of algorithmic problems, P and NP. Decision problems, localization and optimization. Description of some NP-hard problem.  Definition of AI, intelligent machine. Difference between operating systems and intelligent systems. Review of knowledge representation, review of AI languages​​. Applications in industry and services (robotics, planning, waste management). Turing Test.  Hours / week (s)  10 hrs / 2 weeks

UNIT 2: Search State and its Application to Human Games - Machine.

Achievement  The student at the end of the unit will represent problems of artificial intelligence search state known search methods of state to solve, and design human-machine games, which will be practically invincible by man, for it defines the function of more appropriate assessment and will implement various intelligent decision criteria.  Agenda  The problem of search in a state space. Representing and reasoning problems of optimization. Representation of games human problems - machine. Blind Methods. Methods Informed. Evaluation function and its application in computer-human machine. Algorithms human game machine. Decision criteria for human games - Equipment: non-deterministic, greedy, min-max, unlike better profits.  Hours / week (s)  25 h / 5w

UNIT 3: Knowledge Engineering

Achievement  The student will understand at the end of the unit who are expert systems (IS) applications know when to apply, and knowledge-based modeled following the methodology CommonKADS facto systems.  Agenda  Definition of SE. Architecture, taxonomy, applications, and advantages of SE. Requirements for development of SE. Introduction to knowledge engineering. Development methods based on knowledge (SBC) systems. The CommonKADS (CK) methodology. Building a SBC using CK: context model, conceptual model and design model.  Hours / week (s)  10 hrs / 2w

UNIT 4: Development and validation of Expert Systems

Achievement  The student at the end of the unit will design, implement and validate an expert system for it to acquire and learn knowledge representation, inference engines design, calculate performance ratios of an expert system, and reinforce what they learn through various classic examples of systems experts.  Agenda  Acquisition of knowledge. Building the fact base and knowledge base. Knowledge representation structures. The inference engine. The methods of regressive and progressive linkage. Major errors in the development of an SE. Quality of a SE. Efficiency and error SE. Index of agreement, sensitivity and specificity.  Hours / week (s)  15 h / 3w

UNIT 5: Introduction to Machine Learning and Heuristics

Achievement  The student will understand that is the end of combinatorial optimization and machine learning and their applications to create value and increase the competitiveness of organizations through case studies and workshops.  Agenda  Concepts of learning and machine learning. Expert systems vs machine learning. Learning techniques and stages of development of machine learning. Applications of machine learning in industry and services. The combinatorial optimization problem. Concepts of heuristics and meta-heuristics. Exact algorithms vs heuristic algorithms. Heuristics and meta-heuristics. Applications of combinatorial optimization in industry and services.  Hours / week (s)  5 h / 1w

5. CONTENT FOR WEEKS: 

WEEK	CONTENT	CLASSES	LABORATORY WORKS
1°	Classification of algorithmic problems  Presentation of the course. Classification of algorithmic problems, P and NP problem. Decision problems, localization and optimization. Description of some NP-hard problem.  References: [4] Chapter 1 [1] Annex A.	Class Week 1 Summary Week 1
2°	Fundamentals of Artificial Intelligence  Definition of Artificial Intelligence. Intelligent machine. Difference between systems (robotics, planning, waste management). Turing Test.  References: [1] Chapter 1, [2] Chapter 1 [9], Chapter 1.	Class Week 2 Summary Week 2	Examples of Intelligent Systems Exercises LISPWORKS
3° y 4°	Representation problems human game - search machine as a state space  Definition of AI problems as search problems in a state space. Representation of games human problems - machine.  References: [1] Chapters 3, [3] Chapter 2 [4] Chapter 3.	Class Week 3-4 Summary Week 3-4	Binary Tree Exercises functions in LISP
5°	Blind search methods  The problem of search in a state space as a graph. The impossibility of using shortest path methods. The concept of the methods of blind and informed search. The state tree. Blind methods: breadth, depth, non-deterministic.  References: [1] Chapter 4 [2], Chapter 5, [3] Chapter 3, [4] Chapters 5, [9] Chapters 9.	Class Week 5 Summary Week 5	Problem of fox, chicken and corn Searches for depth and breadth in LISP
6°	Search methods reported  The evaluation function, methods that use additional information: first, the best, climb the hill, A *, branch and bound.  References: [1] Chapter 4 [2], Chapter 5, [3] Chapter 3, [4] Chapters 5, [9] Chapters 9.	Class Week 6 Summary Week 6	Avara Search (with routes of Mexico)
7°	Search methods for human games - machine  Algorithm human game - machine. Strategy game machine: non-deterministic, best-first, min-max difference and better profits. Min-max algorithm and alpha-beta.  References: [1] Chapter 5 [2], Chapter 6, [3] Chapter 4 [4] Chapters 6, [9] Chapters 12.	Class Week 7 Summary Week 7	Solution Qualified Practice
8°	Partial Exam
9°	Fundamentals of expert systems  Definition of Expert Systems. Architecture of an expert system. Taxonomy and expert systems applications. Requirements for the development of expert systems and advantages of the use of expert systems. Some problems based on knowledge.  References: [6] Chapter 1.	Week 9
10°	Knowledge Engineering  Introduction. Acquisition of knowledge. C ommonKADS methodology. Design of Expert Systems (ES). Lifecycle of a SE.  References: [6] Chapters 6, [7] Chapter 19.	Week 10
11°	Knowledge Acquisition  Acquisition of knowledge. Building the fact base and knowledge base. Structures of knowledge representation (rules of inference, frames, objects, ontologies, metadata, thesaurus).  References: [6] Chapters 6, [7] Chapter 19.	Week 11	Neural Network - Example
12°	Developing rules based expert systems  Building the fact base and knowledge base. The inference engine. Methods regressive, progressive and reversible chain. Matching techniques, the RETE algorithm. Conflict resolution techniques.  References: [1] Chapters 6 and 8, [2] Chapter 7 [6] Chapter 3 [7], Chapter 3.	Week 12	Neural network with 4 variables
13°	Quality and validation of expert systems  Major errors in the development of an expert system. Quality of an expert system. Validation of intelligent systems, quantitative validation methods. Efficiency and expert systems error. Review the functionality of SE 2nd job.  Tasks: quality and validation exercises SE, validate the proposed system 2nd job.  References: [4], [7] Chapter 21.	Week 13	How to implement an expert system? Expert system
14°	Introduction to Machine Learning and heuristics  Concepts of learning and machine learning. Expert Systems vs machine learning. Learning techniques and stages of development of machine learning. Applications of machine learning in industry and services. The combinatorial optimization problem. Concepts of heuristics and meta-heuristics. Exact algorithms vs heuristic algorithms. Heuristics and meta-heuristics. Combinatorial optimization problems in industry and services.  References: [10], [11], [5] Chapter 1 [8] Chapter 1.	Week 14	Test PROLOG
15°	Introducing Computacional works Students show their skills in the development of expert systems and their applications in the sectors of industry and services. Students will present a report and software.
16°	Final Exam	Previous Final Exams	Final Exam 2014-2

Laboratory 

During the lab sessions the development of computational work, complementary exercises accompany the theory, and develop a programming language for artificial intelligence at basic level LIPS (or a variant thereof) or CLIPS. 

Computer jobs 

Students grouped in 3 will develop two computational work. 

The first work is a set of human-computer interaction with the use of artificial intelligence, the game will be different from the games that exist in the media. The difference between two sets is based: the difference in the environment, differences in rules and objects, or difference in both. The game will use the technique of search in a state space, the human-machine algorithm, and should consider 3 levels of difficulty (beginner, normal and expert). You must have appropriate interfaces and the programming language is free. 

The second work is a rule-based expert system. The knowledge base must be at least 50 rules. You must have appropriate interfaces and the programming language used must be an AI language. Regarding the development should follow the methodology CommonKADS. 

The computational work must be accompanied by a report prepared in the format of an article and be exposed.

6. METHODOLOGY

The course is developed through theoretical - practical activities, emphasizing applications in industry and services. The students were divided into teams of 3 students will develop two computational work. During lab sections CLIPS non procedural language or a variant of LISP will study and advancement of the computational work will be evaluated.

7. EVALUATION  

The average Final (PF) is determined as follows: 

    PF = 0.0125 (CL1 + CL2 + CL3 CL4 +) + 0.01 (TB1 + TB2) + 0.15 + 0.30 * LA * (EA + EB) 

where: 

CLx: Readings controls (CL1, CL2, CL3 and CL4) 

TB1: Group Work (Intelligent Games Man - Machine) 

TB2: Group Work (Expert Systems) 

EA: Partial Exam 

EB: Final Exam 

LA: Laboratory 

Students may substitute exam partial or final if it was unable to provide any of these tests. 

One student will be evaluated to submit 70% or more attendance, this is 5 fouls are automatically withdrawn from the course.

8. REFERENCES

[1] STUART, RUSSELL; PETER, NORVIG

2010 Artificial Intelligence: a modern approach. Ed. Prentice Hall.

ISBN 0-13-103805-2

[2] PATRICK, WINSTON

1984 Inteligencia artificial. Ed. Addison-Wesley

ISBN 0-201-51876-7

[3] ELAINE, RICH

1988 Inteligencia artificial. Ed McGraw-Hill

ISBN 0-07-450364-2

[4] DAVID, MAURICIO

2009Apuntes de inteligencia artificial.

[5] BONIFACIO, MARTIN; ALFREDO, SANZ

2002 Redes neuronales y sistemas difusos. Ed. Alfaomega

ISBN 84-7897-466-0

[6] JOSEPH GIARRATANO – GARY RILEY

2001 Sistemas expertos, principios y programación. Ed. Ciencias Thomson

ISBN 970-686-059-2

[7] JOSÉ PALMA M., ROQUE MARIN M.

2008 Inteligencia artificial, técnicas métodos y aplicaciones. Ed. Mc Graw Hill

ISBN 978-84-484-5618-3

[8] JOSE R. HILERA, VICTOR J. MARTINE.

2000 Redes neuronales artificiales, fundamentos, modelos y aplicaciones. Ed. Alfaomega – rama

ISBN 978-84-484-5618-3

[9] NILS J. NILSON

2001 Inteligencia artificial, una nueva síntesis. Ed. Mc Graw Hill

ISBN 978-84-484-5618-3

[10] CAMPELO Ruy; MACULAN Nelson.

      1994, Algoritmos e Heurísticas. Ed. Universidad Federal Fluminense.

GLOVER Fred; KOCHENBERGER Gary A.

      [11] 2003 HandBook of Metaheuristic. Kluwer International Series.