Module Code:
SD3012
Module Name
Artificial Intelligence
Module Level
3
Semester
B
Issue Date
15 April 2014
Submission Date:
8th May 2014
Weighting
50%
Main Aims of the Module
I. Representation and reasoning paradigms used in AI in both theory and practice with careful attention to the underlying principles of logic, search, and probability.
II. Introduction to the underlying issues in cognitive emulation and to provide an opportunity for practical exercises in logic and probability.
Learning Outcomes for the Module
At the end of this Module students will be able to
· Demonstrate an understanding of search, logic based knowledge representation of issues in planning and learning.
· Be critically aware of the limitations of current symbolic AI paradigms.
· Develop and select appropriate search paradigms for advanced problems.
· Evaluate knowledge of Bayes' Rule and its use in Belief Networks and be able to solve problems concerning the updating of prior probabilities, and to construct belief networks for simple problems.
· Design and evaluate a simple agent system and associated ontology.
· Design, develop and implement a forward chaining knowledge based system including both ontology and rule based using a formalism such as CLIPS.
Assignment
To study intelligent systems and develop an algorithm for an intelligent system.
In this coursework you are required to study intelligent systems and develop an algorithm for an AI Agent based on forward chaining knowledge based system. You are required to identify a suitable ‘domain’ of your choice, acquire domain knowledge and convert the domain knowledge into a suitable rule structure. You should develop your algorithm by using appropriate knowledge representation and searching techniques used in the field of Artificial Intelligence. Your solution should demonstrate the application of Baye’s Network in establishing conditional dependency of the variables/parameters to handle uncertainties. The algorithm must be able to solve a given problem based on the facts and the rules stored in the working memory as a knowledge base for the given system. Evaluate the outcome of your algorithm by running at least two instances, formalized in CLIPS, through the system. The solution of the problem should be interpreted in the human language.
Project Report
Every student will be required to submit a report of 2000 words including the following information
I. A cover page with your name, your student ID, module title and code and the name of the project
II. Details of the work mentioned above.
III. As a conclusion, you will discuss the advantages and the drawbacks of your solution, and possible extensions for the application.
Note: No coursework would be accepted without the Turnitin report. (Maximum accepted level of similarity is 20%). The quotations referenced in the research should be taken from a range of authentic sources.
Generic assessment criteria
The be ...
2. Main Aims of the Module
I. Representation and reasoning paradigms used in AI in both
theory and practice with careful attention to the underlying
principles of logic, search, and probability.
II. Introduction to the underlying issues in cognitive emulation
and to provide an opportunity for practical exercises in logic
and probability.
Learning Outcomes for the Module
At the end of this Module students will be able to
· Demonstrate an understanding of search, logic based
knowledge representation of issues in planning and learning.
· Be critically aware of the limitations of current symbolic AI
paradigms.
· Develop and select appropriate search paradigms for advanced
problems.
· Evaluate knowledge of Bayes' Rule and its use in Belief
Networks and be able to solve problems concerning the
updating of prior probabilities, and to construct belief networks
for simple problems.
· Design and evaluate a simple agent system and associated
ontology.
· Design, develop and implement a forward chaining knowledge
3. based system including both ontology and rule based using a
formalism such as CLIPS.
Assignment
To study intelligent systems and develop an algorithm for an
intelligent system.
In this coursework you are required to study intelligent systems
and develop an algorithm for an AI Agent based on forward
chaining knowledge based system. You are required to identify
a suitable ‘domain’ of your choice, acquire domain knowledge
and convert the domain knowledge into a suitable rule structure.
You should develop your algorithm by using appropriate
knowledge representation and searching techniques used in the
field of Artificial Intelligence. Your solution should
demonstrate the application of Baye’s Network in establishing
conditional dependency of the variables/parameters to handle
uncertainties. The algorithm must be able to solve a given
problem based on the facts and the rules stored in the working
memory as a knowledge base for the given system. Evaluate the
outcome of your algorithm by running at least two instances,
formalized in CLIPS, through the system. The solution of the
problem should be interpreted in the human language.
Project Report
Every student will be required to submit a report of 2000 words
including the following information
I. A cover page with your name, your student ID, module title
and code and the name of the project
II. Details of the work mentioned above.
III. As a conclusion, you will discuss the advantages and the
drawbacks of your solution, and possible extensions for the
4. application.
Note: No coursework would be accepted without the Turnitin
report. (Maximum accepted level of similarity is 20%). The
quotations referenced in the research should be taken from a
range of authentic sources.
Generic assessment criteria
The best grades will be awarded for the student who will
demonstrates a breadth and depth of substantive knowledge that
is exceptional and informed by the highest level of scholarship.
There will be evidence of excellent integration of the full range
of appropriate principles, theories, evidences and techniques.
The student should be able to demonstrate sound judgment
within the given parameters and analyse the question critically.
The student should also demonstrate originality in the
application of the knowledge and should use correct, relevant
and preferably latest reference in support of their findings.
The report should be written in correct and fluent English
language using recommended format with clear aims and
objectives and a summary of the level of achievement of the
objectives.
Note: The further a student’s work diverges from the ideal
described above, the lower their resulting grade is likely to be.
Detailed Marking Criteria
Learning Outcome
70% + Marks
60%-70% Marks
50%-60% Marks
35%-50% Marks
Below 35% Marks
5. Marks (Total 100)
· Full bibliography and appropriate referencing
· Accurate and concise use of sources
· Bibliography provided with omissions in referencing
· Mostly accurate and concise use of sources
· Satisfactory Bibliography with no major omissions in
referencing
·
· A very limited bibliography and an inappropriate referencing
· Limited references to reading;
· Inadequate and incorrect referencing and bibliography
· Poor or little referencing and irrelevant sources
· A wide range of recent, relevant and appropriate reading;
· Strictly written in the recommended format and of an
appropriate length;
· A range of recent, relevant and appropriate reading but not as
wide, recent or relevant as required to achieve all the learning
outcomes.
· Written in the recommended format and of an appropriate
length with minor variations
· Satisfactory level of reading of recent and relevant material
· Overall written in standard format with appropriate style and
length
· Showing variance from the recommended format and may not
follow appropriate length
6. · Poor presentation, language mistakes and not in academic
style
· Not following the permissible length of the work
Demonstrate an understanding of search, logic based knowledge
representation of issues in planning and learning.
A complete overview of the topic with clear background of the
aim, objectives and methodology.
Excellent and a systematic understanding of key aspects of
search, logic and knowledge representation.
A complete overview of the topic with a good attempt to present
background, aim and objectives of the study.
A good understanding of key aspects of search, logic and
knowledge representation
Overview of the topic is presented but lacks in clarity in terms
of the final outcome of the study and the methodology.
Satisfactory understanding of key aspects of search, logic and
knowledge representation
Limited understanding of key aspects of search, logic and
knowledge representation with unclear and inadequate overview
of the topic
Poor overview of the topic presented with little or no attempt to
explain the aim, objectives and methodology of the study.
Limited understanding of key aspects of search, logic and
knowledge representation.
25
Be critically aware of the limitations of current symbolic AI
paradigms
7. Excellent ability to identify and comment upon AI paradigms
with a full awareness of limitations of these paradigms
A good ability to identify and comment upon the AI paradigms
showing good awareness of limitations of these paradigms
Satisfactory ability to identify and comment upon the AI
paradigms showing good awareness of limitations of these
paradigms
Limited ability to describe and comment upon AI paradigms
demonstrating a lack of awareness of limitations of these
paradigms
Poor or no ability to describe or comment upon AI paradigms
7
Develop and select appropriate search paradigms for advanced
problems
Identification of parameters and use of search techniques
appropriate to develop the algorithm and determine its outcome.
Some omissions in identification of parameters and use of
search techniques appropriate to develop the algorithm and
determine its outcome
Some omissions in identification of parameters and use of
search techniques resulting in some errors in the algorithm
outcome
Inaccuracies and major omissions in identification of
parameters and use of search techniques resulting in some
sizeable.
Limited or no ability shown in the identification of parameters
and use of search techniques appropriate to develop the
algorithm and determine its outcome
13
Evaluate knowledge of Bayes' Rule and its use in Belief
Networks and be able to solve problems concerning the
updating of prior probabilities with and to construct belief
networks for simple problems.
8. Excellent integration of the full range of appropriate rules and
methods to solve the problem
A good integration of the full range of appropriate rules and
methods to solve the problem
Satisfactory integration of a range of appropriate rules and
methods to solve the problem
Limited integration of a range of appropriate rules and
methods to solve the problem
Fails to describe and integrate appropriate rules and methods to
solve the problem.
12
Design and evaluate a simple agent system and associated
ontology.
Excellent understanding of different blocks and functioning of
an intelligent system.
Able to critically evaluate the system design.
Correct logic, correct formation of the rules and completely
functional algorithm.
A good understanding of different blocks and functioning of an
intelligent system.
Some shortcomings in critical analysis of the system design
Some inconsistencies in the logic and some mistakes in the
rules.
Satisfactory understanding of different blocks and functioning
of an intelligent system
9. Some shortcomings in the critical analysis of the system design.
Some errors in the logic and rule formation..
Limited understanding of different blocks and functioning of
an intelligent system
Major shortcomings in the critical analysis of the system
design.
Major errors in the logic and rule formation. Algorithm not
working as required. Algorithm not working as required
Inadequate research and analysis of different blocks and
functioning of an intelligent system.
No or limited ability shown to develop a working algorithm.
25
Design, develop and implement a forward chaining knowledge
based system including both ontology and rule based using a
formalism such as CLIPS
Excellent ability to employ accurately forward chaining rules in
the development of the algorithm
Good ability to employ forward chaining rules with minor
inaccuracies
Demonstrate satisfactory knowledge of forward chaining system
but with a limited ability to employ it accurately
Limited knowledge of forward chaining system and a limited
ability to employ it accurately
Limited knowledge of forward chaining system and no ability
to employ the rules for practical purposes
18
Total
100