Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 24

CASE BASED REASONING

YUVARAJ G
09MCS107
TECHNICAL SEMINAR-II
OUTLINE
• INTRODUCTION
• CASE BASED REASONING
• CBR PROCESS
• CBR CYCLE
• CASE BASE
• CASE BASE ORGANIZATION
• APPLICATION
• ADVANTAGE
• DISADVANTAGE
• REFERENCE
INTRODUCTION
• To solve a new problem by remembering a
previous similar situation and by reusing
information and knowledge of that situation.
• Some typical problem solving situations:
– An auto mechanic who fixes an engine by recalling
another car that exhibited similar symptoms
– A physician uses the diagnosis and treatment of the
previous patient to determine the disease and
treatment for the patient in front of him
CASE BASED REASONING
• A methodology to model human reasoning and thinking
• A methodology for building intelligent computer systems
• CBR store previous experience (cases) in memory
• To solve new problems:
– Retrieve similar experience about similar situations from the
memory
– Reuse the experience in the context of the new situation
complete or partial reuse, or adapt according to differences
– Store new experience in memory (learning)
CBR PROCESS
• CBR terminology
– Case denotes a problem situation
– Past case denotes previously experienced situation
– New case or Unsolved case denote a new problem to
be solved
• CBR Cycle consist of four process
– RETRIEVE
– REUSE
– REVISE
– RETAIN
CBR CYCLE
CASE BASE
• Case base represents set of cases.
• Each case describes one particular diagnostic
situation
• Each case records several features and their
specific values occurred in that situation
• Typically case comprises
– Problem statement
– Solution
– Outcome
– Facts and supporting reasoning
CASE BASE
CASE BASE ORGANIZATION
• Case base should be organized into manageable
structure that support
– Accurate search and retrieval
– Efficient search and retrieval
• Approaches to organization
– Flat organization
• Simple
• Time consuming, inefficient CBR system
– Clustered organization
• Similar case are grouped
• Require complex algorithm
– Hierarchical organization
• cases that share the same features are grouped together
• Expensive case base reorganization
RETRIEVE
• The Retrieve task starts with a (partial)
problem description, and ends when a
best matching previous case
• Involves following activities
– IDENTIFY FEATURE – comes up with a set
of relevant problem descriptors
– MATCHING – returns a case sufficiently
similar to the new case
– SELECT – chooses the best match
SOLVING A NEW CASE
RETRIEVAL TECHNIQUE
• Nearest-Neighbor Retrieval
– computes the similarity between stored cases and new input case based
on weight features.
– Simple
– Slow retrieval speed when the case base is large

• Inductive Retrieval
– determine feature that discriminate cases and generates a decision tree
type structure to organize the cases
– approach is very useful when a single case feature is required as a
solution
– Fast case retrieval
– Time consuming and impossible to retrieve missing or known case data
NEAREST-NEIGHBOR RETRIEVAL
REUSE
• Past case retrieved through RETRIEVE
can reuse either by
– Applying some solution to new case
– Adapt the required part of past case and
apply to new case
• To main way to adapt past cases
– Structural adaptation
– Derivational adaptation
REUSE
REVISE
• Consist of two task
• Evaluate case solution
– Applying the solution in the real environment
– Results may take some time to appear depending on
the type of application
– Task suggest additional adaptation or repairs
requirement for proposed solution
• Repair fault
– Uses feedback of evaluation phase to over come
failure
– Revised plan can then be retained directly if corrected
– Otherwise it can be evaluated and repaired again
RETAIN
• Case base is updated with a new case
• Indexing and integrating case into case
organization and used for later retrieval
APPLICATION FIELDS
CLAVIER
• Problem- how to optimize the loading of an
autoclave for curing aircraft components during
manufacture
• CLAVIER is developed with following objective
– Reuse previous successful configuration
– Reduce the work load of the expert
– Train new personnel
• Cases contains
– Parts and their relative positions in a table
– Tables and their relative position in the autoclave
– Parameters such as start and finish time, pressure
and temperature
CHEF- A MEAL PLANNER
• CHEF accepts goals that specify different
types, taste and texture of dishes
• Input is of the form
– Create a meat dish
– Include green beans
– Use boiling
– Achieve a crisp texture
• Output is a single recipe that satisfies the
constraint
CHEF STEPS
• Retrieve an old recipe that fulfills as many
of its new goals
• Adapt the old plan, to fit the new situation
– Reinstantiates the old plan
– Modify the old plan for the new situation
• Use the plan and collect feedback about
working of the plan.
– If it worked properly then store for reuse
– If not repair the faulty plan
ADVANTAGE
• Reduces the knowledge acquisition effort
• Requires less maintenance effort
• Improve problem solving performance
through reuse
• Makes use of existing data, e.g. in
databases
• Improve over time and adapt to changes in
the environment
• High user acceptance
DRAWBACK
• High memory/storage requirements and
time-consuming retrieval
• Difficulties in handling dynamic problem
domains
• Problem in handling noisy data
• Some problem situations can occur for
which the system has no solution
REFERENCE
• Peter Jackson, “ Introduction to Expert System”, Pearson
Education, 2003
• Rajendra Akerkar, “Introduction to Artificial Intelligence”,
PHI Learning Private Limite,2009
• A. Aamodt, E. Plaza, ”Case-Based Reasoning:
Foundational Issues, Methodological Variations, and
System Approaches,” AI Communications. IOS Press,
Vol. 7: 1, pp. 39-59.
• R. Bergmann,” Introduction to Case-Based Reasoning”,
University of Kaiserslautern
• http://groups.yahoo.com/group/case-based-reasoning/
• http://en.wikipedia.org/wiki/Case-based_reasoning

You might also like