Artificial Intelligence

Instance: 2022/2023 - 2S

Cycles of Study/Courses

Last updated on 2023-02-06.

Fields changed: Objectives, Bibliografia Complementar, Pre_requisitos, Fórmula de cálculo da classificação final, Provas e trabalhos especiais, Avaliação especial, Melhoria de classificação, Obtenção de frequência, Programa, Tipo de avaliação, Lingua de trabalho, Software de apoio à Unidade Curricular, Componentes de Avaliação e Ocupação, Bibliografia Obrigatória, Bibliografia Complementar, Objetivos, Resultados de aprendizagem e competências, Pre_requisitos, Fórmula de cálculo da classificação final, Provas e trabalhos especiais, Avaliação especial, Melhoria de classificação, Obtenção de frequência, Programa, Tipo de avaliação, Lingua de trabalho, Software de apoio à Unidade Curricular, Componentes de Avaliação e Ocupação, Bibliografia Obrigatória, Resultados de aprendizagem e competências

Teaching language

Objectives

This course provides a set of subjects (topics) that are the core of the Artificial Intelligence and Intelligent System area. The main objectives are:
1. Understand the fundamentals of Artificial Intelligence and Intelligent Systems, what characterizes and distinguishes them and their applicability.
2. Being able to design and implement Agents and Multi-Agent Systems to solve different problems.
3. To learn heuristic and systematic methods of problem solving, with and without adversaries and optimization algorithms.
4. To learn methods of acquisition, representation and reasoning with uncertain knowledge using different formalisms.
5. To understand the basis of natural language processing and its applications.
6. Know and be able to apply learning algorithms with different paradigms (supervised, unsupervised, reinforcement, evolutionary, deep learning) and algorithms (decision trees, neural networks, SVMs).
7. To understand advanced topics in Artificial Intelligence and be able to formulate a vision into the future of AI.
8. To develop simple but complete projects using AI techniques.

Percentual Distribution: Scientific component: 50%; Technological component: 50%

Learning outcomes and competences

At the end of the course, students should be able to represent, acquire, manipulate and apply knowledge using computer systems. More specifically, the student should be able to:

• Understand the fundamentals of Artificial Intelligence and Intelligent Systems.
• Understand the notion of Computational Agent and Multi-Agent System and be able to design and implement Agents and Multi-Agent Systems to solve different problems.
• Be able to apply search and optimisation methods and algorithms to solve complex problems with and without Adversaries.
• To learn methods of Representation and Reasoning with uncertain Knowledge using different formalisms.
• Be able to apply machine learning algorithms with different paradigms (Induction, Evolutionary, Reinforcement, Neuronal, Deep Learning).
• Understand advanced topics in Artificial Intelligence and be able to formulate a vision into the future of AI and its practical applications in the present and in the future.

Working method

Pre-requirements (prior knowledge) and co-requirements (common knowledge)

Knowledge of algorithms, data structures, and programming.

Program

Definition of AI. Fundamentals, Scope, Evolution and Chronology of AI. Problems and Approaches of AI and Intelligent Systems. AI Applications.

The Concept of Agent. Environments. Agent Architectures: Reactive, Deliberative, Goal-Based, Utility-Based, Learning and BDI. Multi-Agent Systems: Concept, Motivation, Architectures, Communication, Coordination. Practical Examples of Application.

III. Problem Solving Methods

Problem Formulation. State Space. Search Strategy. Uninformed Search: Breadth First, Depth First, Uniform Cost, Iterative Deepening, Bidirectional Research. Intelligent Search: Greedy Search, A* Algorithm. Search with Adversaries: Game Search, Minimax Algorithm, Alpha-Beta Cuts, Monte Carlo Tree Search, Search with Imperfect Information. Practical Examples of Application.

Formulation of Decision/Optimization Problems. Hill-Climbing Algorithm, Simulated Annealing, Tabu Search, "Ant Colony". Genetic Algorithms and Evolutionary Computation. Constraint Satisfaction. Practical Examples of Application.

Knowledge Representation and Reasoning. Propositional and Predicate Logic. Semantic Networks, Frames, Rules, and Ontologies. Logic Programming and Programming with Constraints. Reasoning with Uncertain Knowledge. Knowledge-Based Systems. Expert Systems. Practical Examples of Application.

History and Motivation for Machine Learning. Main Types of Machine Learning: Supervised Learning, Unsupervised Learning and Reinforcement Learning. Deep Learning Concept. Applications of Machine Learning. Data: Types, Data Quality, Preprocessing and Transformation. Model Interpretation and Evaluation. ML Tools, and Libraries. Algorithms: Decision Trees. K-Nearest Neighbour. Artificial Neural Networks, Support Vector Machines. Practical Application Examples.
Reinforcement Learning: State, Action, Policy, Reward and Value. Exploration-Exploitation Tradeoff. Markov Decision Processes. Tools and Libraries. Algorithms: Qlearning, SARSA, DQN, PPO and SAC. Practical Application Examples.

VII. Natural Language Processing

Introduction to NLP. Levels of Processing. Classical approach. Grammars with Defined Clauses. Statistical Approach. Text Mining. NLP Tasks. Languages Resources. NLP Applications. Machine Learning in NLP. Basic Text Processing: Normalization, Tokenization, Segmentation. Text Classification. Bag of Words. Naive Bayes. Generative vs Discriminative Classifiers. Word Embeddings. Deep Learning in NLP. Practical Application Examples.

VIII. Advanced Topics in Artificial Intelligence

Perception/Vision, Communication, Interaction, Planning, Scheduling, Robotics, Intelligent Simulation, Social Intelligence. Applications of Artificial Intelligence and Intelligent Systems. The Future of AI. IA and the Society. Beneficial IA. Explainable AI. Machine Ethics. Weak and Strong IA. Super Intelligence. The Singularity.

Mandatory literature

Teaching methods and learning activities

Theoretical classes: exposition with interaction. Theoretical-practical classes: modeling, problem-solving, programming exercises and project development.

keywords

Evaluation Type

Assessment Components

Amount of time allocated to each course unit

Eligibility for exams

Not exceed the absence limit allowed and have a minimum of 7.5 (out of 20), in the Practical Work Component (PW)

Calculation formula of final grade

PW = Practical Work, MT = Mini-Tests, PC - Participation in Class, EA = Exam of Appeal.

Final Classification = 40%*PW + 40%*MT + 20%*PC

PW Component includes two practical works, each with 50% weight.
MT Component includes two mini-tests, each with weight of 50%.
PA Component includes participation in the classes and realization of Kahoots.

Pass mark of 7.5 (in 20) in MT component needed for approval. 

Appeal Classification = 40%*PW+ 60%*EA

Examinations or Special Assignments

Internship work/project

Special assessment (TE, DA, ...)

The evaluation rules apply to all students, regardless of their status. Students who, by their status, are excused from attending practical classes should contact the teachers for monitoring the work.

Classification improvement

The minitests are improvable at the time of appeal. The practical part is not improvable.