Memory and Experience-Based Learning

You are here

Home » Studies » Masters » Master in Artificial Intelligence » Curriculum » Syllabus » Memory and Experience-Based Learning
Credits
4.5
Types
Elective
Requirements
This subject has not requirements, but it has got previous capacities
Department
CS
The course is devoted to Memory-based Experiential Learning, focusing in a symbolic cognitive approach: Case-Based Reasoning. The basic components of the CBR reasoning cycle will be carefully studied and analysed. The application of CBR to the real world will be reviewed, and main problems the deployment of CBR application will be outlined.

Teachers

Person in charge

  • Javier Vazquez Salceda ( )
  • Ramon Sangüesa Sole ( )

Weekly hours

Theory
1.3
Problems
0
Laboratory
0.8
Guided learning
0.1
Autonomous learning
5.3

Competences

Generic Technical Competences

Generic

  • CG1 - Capability to plan, design and implement products, processes, services and facilities in all areas of Artificial Intelligence.

Technical Competences of each Specialization

Academic

  • CEA3 - Capability to understand the basic operation principles of Machine Learning main techniques, and to know how to use on the environment of an intelligent system or service.
  • CEA12 - Capability to understand the advanced techniques of Knowledge Engineering, Machine Learning and Decision Support Systems, and to know how to design, implement and apply these techniques in the development of intelligent applications, services or systems.

Professional

  • CEP2 - Capability to solve the decision making problems from different organizations, integrating intelligent tools.
  • CEP5 - Capability to design new tools and new techniques of Artificial Intelligence in professional practice.

Transversal Competences

Teamwork

  • CT3 - Ability to work as a member of an interdisciplinary team, as a normal member or performing direction tasks, in order to develop projects with pragmatism and sense of responsibility, making commitments taking into account the available resources.

Analisis y sintesis

  • CT7 - Capability to analyze and solve complex technical problems.

Basic

  • CB8 - Capability to communicate their conclusions, and the knowledge and rationale underpinning these, to both skilled and unskilled public in a clear and unambiguous way.

Objectives

  1. Being able to decide when a problem is suitable to be solved through an experiential learning scheme (a CBR paradigm)
    Related competences: CEA3, CEA12, CT7,
  2. To be able to design a CBR case structure (problem description, solution) for a given realistic problem
    Related competences: CEA3, CEA12, CG1, CEP2, CT3, CT7,
  3. To be able to design and implement a Case Library (selecting the proper indexing mechanisms, library structure and similarity functions) for a given realistic problem
    Related competences: CEA3, CEA12, CG1, CEP2, CEP5, CT3, CT7,
  4. To be able to design and implement an appropriate adaptation function (adapting solutions from previous cases to a new one) for a given realistic problem
    Related competences: CEA3, CEA12, CG1, CEP2, CEP5, CT3, CT7,
  5. To be able to design and implement a CBR Maintenance mechanism (defining a case relevance metric, selecting a maintenance strategy, implementing a library maintenance module) for a given realistic problem.
    Related competences: CEA3, CEA12, CG1, CEP2, CEP5, CT3, CT7,
  6. To be able to validate a CBR prototype (create a set of case examples, validate all CBR components) and analise the results.
    Related competences: CEA3, CEA12, CG1, CEP2, CT3, CT7, CB8,
  7. Get some basic knowledge on Cognitive AI theories and methods for Memory-based learning (Exemplar Learning, Instance-based Learning, Experiential Learning, Case-Based Learning) and their foundations on Cognitive Sciences.
    Related competences: CEA3, CEA12, CG1, CT7,

Contents

  1. Human memory theories and their relevance to AI
    Basic overview of the role of memory in learning principles and classification of Machine Learning techniques
  2. Memory and learning in cognitive AI
    Early AI and symbolic systems are examined, focusing on the physical symbol system hypothesis and early views on learning and memory as symbol manipulation and retrieval.

    Conceptual analysis of classic AI programs to identify implicit cognitive assumptions about memory
  3. Exemplar and instance theories of learning
    Models based on examples and instance theories from cognitive psychology are presented as alternatives to rule-based abstraction, emphasizing similarity-based generalization and episodic traces.
    How example theories challenge classical symbolic perspectives on learning is analyzed.
  4. Cognitive foundations of instance-based learning
    Cognitive Foundations of Instance-Based Learning

    Instance-Based Learning (IBL) is presented as a computational analogue of example-based cognition, framing learning as memory accumulation rather than model induction.

    Reasoning with explicit instances and similarity judgments in small decision problems.
  5. Algorithmic structure of instance-based learning
    Algorithmic structure of learning based on instances

    Formal IBL algorithms, more accurate research methods, similarity metrics, incremental learning and employment policies.
  6. Overview of techniques and limitations in Exemplar Learning and IBL
    Scalability, noise sensitivity, feature relevance, and the need for knowledge-based memory control.
  7. Experiential Learning
    Experience and episodes. Experiential learning.
  8. CBR System Components
    Description and analysis of the basic components, architecture and processes of CBR systems
  9. CBR Academic Demonstrators/Examples
    Review of the most significant CBR systems and comparison of features
  10. CBR Application on a real domain
    A real application will be described and analysed.
  11. Problems in the development of CBR systems
    a. Competence
    b. Space Performance
    c. Time Performance
  12. Reflective Reasoning in CBR
    Case base maintenance techniques as a form of reasoning and learning
  13. Hybrid Systems
    Description and analysis of CBR neurosymbolic systems
  14. CBR Systems' Evaluation
    The various techniques for evaluating the performance and quality of CBR systems will be studied and applied.
  15. Advanced Research Issues in CBR
    a. Temporal CBR
    b. Spatial CBR
    c. Hybrid CBR Systems
    d. Recommender Systems: CBR as a recommendation tool
    e. Agents and CBR
    f. Distributed CBR

Activities

Activity Evaluation act


Memory and Learning in humans and Cognitive AI


Objectives: 1 7
Contents:
Theory
2h
Problems
0h
Laboratory
0h
Guided learning
0h
Autonomous learning
2h

Exemplar and instance theories of learning. Cognitive foundations.


Objectives: 7
Contents:
Theory
2h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
5h

Exemplar Learning and Instance Based Learning: Cognitive foundations, algorythms and techniques.


Objectives: 7
Contents:
Theory
4h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
6h

Experience and episodes. Experiential Learning.


Objectives: 7
Contents:
Theory
2h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
3h

Fundamentals of Case-based Reasoning & Academic Demonstrators


Objectives: 1 2
Contents:
Theory
2h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
6h

CBR System Components. CBR reasoning cycle.


Objectives: 2 3 4
Contents:
Theory
2h
Problems
0h
Laboratory
2h
Guided learning
0h
Autonomous learning
9h

CBR Application on a real domain


Objectives: 2 3 4 5
Contents:
Theory
1h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
1h

CBR Development Problems


Objectives: 2 3 5
Contents:
Theory
1.5h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
8h

Reflective Reasoning in CBR. CBR System Maintenance


Objectives: 5
Contents:
Theory
1h
Problems
0h
Laboratory
3h
Guided learning
0h
Autonomous learning
7.5h

CBR Applications and Development Tools & CBR Systems' Evaluation


Objectives: 6
Contents:
Theory
1h
Problems
0h
Laboratory
1h
Guided learning
0h
Autonomous learning
7h

Advanced CBR systems Research Issues


Objectives: 2 3 4 5
Contents:
Theory
1h
Problems
0h
Laboratory
0h
Guided learning
0h
Autonomous learning
4h

Practical work (PW) public presentation & discussion


Objectives: 2 3 4 5 6
Week: 15 (Outside class hours)
Theory
0h
Problems
0h
Laboratory
0h
Guided learning
1.5h
Autonomous learning
21h

Teaching methodology

The teaching methodology will include both theoretical lecture sessions, sessions with practical examples of the concepts and algorithms explained in the course, and also some sessions devoted to support the practical work of the students.

The teamgroup work will consist on the design, implementation, application and validation of a Case-Based Reasoning project to solve a realistic problem. The project will be developed in parallel to the topics presented in the course following the structure:
- new topics/techniques are introduced in the classroom
- if these topics/techniques are suitable to be used in the Practical Work, then students are asked to attemp their application as autonomous work.
- the work done is discussed and validated by the lecturer next weeks in the classroom.

Evaluation methodology

Evaluation of the knowledge and skills obtained by the students will be assessed through one practical project work (PW) which will be done on a team group basis.

The teamgroup work will consist on the design, implementation, application and validation of a Case-Based Reasoning project to solve a realistic problem. The project will be developed in parallel to the topics presented in the course following the structure:
- new topics/techniques are introduced in the classroom
- if these topics/techniques are suitable to be used in the Practical Work, then students are asked to attemp their application as autonomous work.
- the work done is discussed and validated by the lecturer next weeks in the classroom.

The final grade will be computed as follows:

FinalGrade= PWGr * WFstud, where 0 <= WFstud <= 1.2

WFstud is a Working Factor evaluating the work of a particular student within his/her teamwork in PW. It will be obtained by observing and assessing the load of work and degree of participation of each student throughout the development of the PW. In normal conditions, the WFstud = 1.


The PWGr will be computed as follows:

PWGr = 0.5 * TeachA + 0.5 * SelfA

where TeachA is the teacher assessment of the teamwork evaluated according to:

- The methodology of the work (0.5)
- The quality of the report written (0.2)
- The quality of the oral exposition (both presentation and content assessed, as well as the ability to answer questions) (0.2)
- Planning, coordination and management of the team (0.1)

and SelfA is the individual assessment of each student by all the members of his/her team.

Bibliography

Basic:

Complementary:

Previous capacities

The requirements are the ones provided by the mandatory courses of the Master, especially those provided by Introduction to Machine Learning (IML).