|
Artificial Intelligence
Scholar Year: 2019/2020 - 1S
| Code: |
INF32185 |
|
Acronym: |
IA |
| Scientific Fields: |
Informática |
Courses
| Acronym |
N. of students |
Study plan |
Curricular year |
ECTS |
Contact time |
Total Time |
| INF |
126 |
|
3º |
6,0 |
90 |
162,0 |
Teaching language
Portuguese
Intended learning outcomes (Knowledges, skills and competencies to be developed by the students)
Give students knowledge of problem-solving methods based on Artificial Intelligence techniques, typically resorting to the representation and use of knowledge. It is intended to provide students with the understanding and programming capability of state-space search algorithms and others used in game theory.
Syllabus
1. Introduction to Artificial Intelligence
1.1. Types of problems and solutions
1.2. Sub-areas of Artificial Intelligence
2. Deepening of LISP as programming languages for Artificial Intelligence
2.1. Recursion and functional programming
2.2. Atoms and Lists: data structures and functions; functions.
2.3. The LISP Evaluator; Meta-functions
3. Troubleshooting
3.1. Definition and characteristics of problems
3.1.1. Combustive explosion
3.1.2. The role of knowledge
3.2. Search in Space of States
3.2.1. Comprehensive Methods
3.2.2. Satisfaction of constraints
3.2.3. Informed methods; heuristics; algorithms
4. Knowledge engineering
4.1. Techniques of knowledge representation
4.1.1. Rules-based systems
4.1.2. Representation of uncertain / incomplete knowledge
4.2. Inference Processes
4.2.1. Inference based on forward chaining and backward chaining
4.2.2. The RETE algorithm
4.3. Methodologies for Developing Expert Systems
5. Theory of games
5.1. Games like state space search problems
5.2. The minimax algorithm
5.3. The alphabet algorithm
Software
LispWorks
Demonstration of the syllabus coherence with the UC intended learning outcomes
Students acquire
- technical skills of functional programming through the teaching of LISP
- ability to work in groups and present their work to a hearing, through the presentation of projects
- ability to solve complex problems by teaching space-state demand
Teaching methodologies
Theoretical classes: presentation with the aid of slides + formative and / or summative evaluation.
Practical classes: learning the necessary concepts of the LISP programming language and solving programming problems in this language + formative and / or summative evaluation.
Lab classes: Programming exercises to solve on the computer. Some follow up in the development of the projects of the discipline + formative and / or summative evaluation. In the laboratories will be used the method of pair-programming including in the resolution of the series of exercises for evaluation.
Demonstration of the teaching methodologies coherence with the curricular unit's intended learning outcomes
The basic concepts of Knowledge Engineering and the development of expert systems will be introduced. In addition, students acquire in this discipline the ability to use a programming language based on a new paradigm, namely the functional paradigm: LISP. To consolidate knowledge about the application of this paradigm will be carried out two projects of application of the theoretical concepts of State Space and Game Theory using the LISP language.
Assessment methodologies and evidences
The evaluation contemplates two possible ways, one of them (B) aimed at students with worker-student status.
Mode A (continuous assessment):
Approval in this Mode is conditional upon completion of at least N-3 sets of exercises in class. Does not contemplate examination.
• Elements:
o Exercises - to be carried out in theoretical, practical and laboratory classes (usually in the last 30 min of the class).
o 2 programming projects (in LISP).
the Test, held in the theoretical class in the last week of classes.
• Final Grade A = Exercises * 30% + Test * 20% + Project1 * 25% + Project2 * 25%
Required to obtain a grade equal to or greater than 7 values (on the 0-20 scale) in each of the assessment elements.
The test can be replaced by the exam (i.e., the best grade) at any evaluation time.
Mode B (evaluation by examination):
Class exercises will not be counted, nor will there be attendance control. Does not contemplate Test. Mode B can be used at any of the evaluation times.
• Elements:
o 2 programming projects (in LISP).
o Final exam.
• Final Grade B = Project1 * 25% + Project2 * 25% + Final Examination * 50%
Required to obtain a grade equal to or greater than 7 values (on the 0-20 scale) in each of the assessment elements
Attendance system
The regular evaluation will have the following components:
1) Series of exercises to be solved during theoretical, practical and laboratory classes. The student may fail to deliver a maximum of 3 grades, or otherwise be required to use Mode B. The grade average will be calculated based on the grades delivered for evaluation.
2) Two programming projects, mandatory for both modes of evaluation.
a) The first project is related to a state space search problem, developed in LISP; The second project will consist of the implementation of a game theory algorithm, in the same language.
b) Projects will be carried out in groups of up to 2 people. The delivery dates of the projects will be indicated in the respective statements, having in both cases 3 days of tolerance with respect to the date marked, discounting one value for each day of delay.
c) In case the student does not submit any of the projects on the scheduled dates in the normal time, or fail, there will be the possibility of delivering a new version of the missing project (s) (slightly different from those of the 1st period but more the same field of application) until the day of the examination of the appeal period, or in the case of special season a project different from the previous ones, until the day of the respective examination. In any case, a tolerance of 3 days will be granted with respect to the date marked, discounting one value for each day of delay.
d) The documentation and the LISP code for the projects will be subject to analysis of plagiarism by appropriate software tools, which means that any indication of copy (between projects of the current school year and also with projects of previous school years) will lead to a penalty on the note of the respective project, and from a certain degree of similarity (not very high) will naturally lead to the cancellation of the project. Situations of extreme plagiarism may lead to disciplinary proceedings.
e) Attention: All projects will be subject to individual oral evaluation, which will include the implementation of LISP code. The oral evaluation will be done after the documentary evaluation of the two projects of each group of students, at a date and time to be marked by the teachers, during the evaluation period defined for the respective time in the school calendar.
Primary Bibliography
Elaine Rich;Inteligencia Artificial |
Secondary Bibliography
Nils Nilsson;Artificial Intelligence |
Stuart Russel and Peter Norvig;Artificial Intelligence: A Modern Approach |
Observations
It is recommended to install and use the free version of the Lispworks development environment, provided by the faculty to follow the practical and laboratory classes.
Hours of doubt: see Moodle
|
|
