Neural Networks and Deep Learning

Credits
6
Types
Compulsory
Requirements
This subject has not requirements , but it has got previous capacities
Department
CS
The goal of machine learning is to develop methods that allow systems to learn from data and improve their performance in the presence of uncertainty. Neural networks and deep learning are key areas of machine learning that use models inspired by the human brain. By organizing artificial neurons into layers, these models can learn patterns and representations from data, while deep learning enables the solution of more complex problems.

This course introduces the fundamental concepts of neural networks and deep learning, including basic architectures, learning algorithms, and practical applications, providing students with the foundations needed to understand and apply these methods.

Teachers

Person in charge

Others

Weekly hours

Theory
2
Problems
0
Laboratory
2
Guided learning
0
Autonomous learning
6

Competences

Transversal Competences

Transversals

  • CT5 [Avaluable] - Solvent use of information resources. Manage the acquisition, structuring, analysis and visualization of data and information in the field of specialty and critically evaluate the results of such management.

    • Basic

  • CB3 - That students have the ability to gather and interpret relevant data (usually within their area of ??study) to make judgments that include a reflection on relevant social, scientific or ethical issues.

    • Technical Competences

    Especifics

  • CE01 - To be able to solve the mathematical problems that may arise in the field of artificial intelligence. Apply knowledge from: algebra, differential and integral calculus and numerical methods; statistics and optimization.
  • CE12 - To master the fundamental principles and models of computing and to know how to apply them in order to interpret, select, assess, model, and create new concepts, theories, uses and technological developments related to artificial intelligence.
  • CE13 - To evaluate the computational complexity of a problem, identify algorithmic strategies that can lead to its resolution and recommend, develop and implement the one that guarantees the best performance in accordance with the established requirements.
  • CE15 - To acquire, formalize and represent human knowledge in a computable form for solving problems through a computer system in any field of application, particularly those related to aspects of computing, perception and performance in intelligent environments or environments.
  • CE18 - To acquire and develop computational learning techniques and to design and implement applications and systems that use them, including those dedicated to the automatic extraction of information and knowledge from large volumes of data.
  • CE20 - To select and put to use techniques of statistical modeling and data analysis, assessing the quality of the models, validating and interpreting.
  • CE26 - To design and apply techniques for processing and analyzing images and computer vision techniques in the area of artificial intelligence and robotics

    • Generic Technical Competences

    Generic

  • CG4 - Reasoning, analyzing reality and designing algorithms and formulations that model it. To identify problems and construct valid algorithmic or mathematical solutions, eventually new, integrating the necessary multidisciplinary knowledge, evaluating different alternatives with a critical spirit, justifying the decisions taken, interpreting and synthesizing the results in the context of the application domain and establishing methodological generalizations based on specific applications.
  • CG8 - Perform an ethical exercise of the profession in all its facets, applying ethical criteria in the design of systems, algorithms, experiments, use of data, in accordance with the ethical systems recommended by national and international organizations, with special emphasis on security, robustness , privacy, transparency, traceability, prevention of bias (race, gender, religion, territory, etc.) and respect for human rights.
  • CG9 - To face new challenges with a broad vision of the possibilities of a professional career in the field of Artificial Intelligence. Develop the activity applying quality criteria and continuous improvement, and act rigorously in professional development. Adapt to organizational or technological changes. Work in situations of lack of information and / or with time and / or resource restrictions.

    • Objectives

    1. To know how to identify a data analysis problem and solve it from start to finish (end to end)
      Related competences: CG4, CG8, CG9, CT5, CE13, CE15,
    2. To know the theoretical foundations of neural networks as models of machine learning
      Related competences: CE26, CG4, CE01, CE12, CE13, CE18, CE20,
    3. To know and understand the fields of application of neural networks and know how to develop solutions to specific problems
      Related competences: CG9, CE12, CE15, CE18,
    4. To know how to design solutions for problems related to language, image or sound
      Related competences: CE26, CG4, CG8, CG9, CT5, CB3, CE13, CE15, CE18,

    Contents

    1. General concepts of machine learning
      Review of the general theoretical concepts of machine learning. Learning as an optimization problem. Bayesian interpretation of the learning problem. Generalized linear models.
    2. Foundations of artificial neural networks.
      Foundations of artificial neural networks. Basic biological concepts. McCulloch-Pitts model. Cognitive and computational implications. Lippmann networks. Loss functions, activation functions.
    3. Feed-forward neural networks
      Feed-forward neural networks.
      Linear networks (I): the Perceptron.
      Linear networks (II): the Delta rule.
      Multilayer Perceptrons and Backpropagation.
      Descent of gradients and variants.
      Other optimizers: pseudo-Newton, CG, Rprop.
      Networks of radial basis functions.
      Autoencoders and VAEs.
      Support vector machines.
      Convolutional networks.
    4. Advanced neural networks
      Hopfield networks.
      Graph neural networks.

    Activities

    Activity Evaluation act


    Theoretical classes

    Development of theoretical classes in the assigned hours. These are eminently masterful classes supported by projections and blackboards.
    Objectives: 1 2 3 4
    Contents:
    Theory
    28h
    Problems
    0h
    Laboratory
    0h
    Guided learning
    0h
    Autonomous learning
    35h

    Laboratory classes

    Examples of the application of the concepts seen in theory classes. Explanations related to the triats programming languages. Additional explanations relevant to the subject: practical skills, experimental methodology, etc.
    Objectives: 1 4
    Contents:
    Theory
    0h
    Problems
    0h
    Laboratory
    28h
    Guided learning
    0h
    Autonomous learning
    25h

    Partial Exam

    Partial exam (in the middle of the semester) that covers all the syllabus seen up to that point, or a little earlier, at the teacher's discretion. The exam will take place in a laboratory classroom and may consist of theory, methodological or practical questions.
    Objectives: 1 2 3
    Week: 9 (Outside class hours)
    Theory
    0h
    Problems
    0h
    Laboratory
    0h
    Guided learning
    0h
    Autonomous learning
    0h

    Final Exam

    Final exam (during the period of final exams) that covers all the syllabus seen in the subject. The exam will be held in a theory classroom and may consist of theory or methodological questions.
    Objectives: 1 2 3 4
    Week: 15 (Outside class hours)
    Theory
    0h
    Problems
    0h
    Laboratory
    0h
    Guided learning
    0h
    Autonomous learning
    0h

    Practical project

    Development of a practical project where you can demonstrate that you know how to apply the concepts, methods and techniques specific to the subject.
    Objectives: 1 4
    Contents:
    Theory
    0h
    Problems
    0h
    Laboratory
    0h
    Guided learning
    0h
    Autonomous learning
    30h

    Teaching methodology

    The course delves into one of the most important machine learning paradigms today: artificial neural networks, with a strong foundation in probability, statistics and mathematics. The theory is introduced in lectures where the teacher explains the concepts. These concepts are put into practice in laboratory classes, where the student learns to develop machine learning solutions to real problems of some complexity. Students must work on and hand in a project at the end of the course.

    Evaluation methodology

    The course is graded as follows:

    P = Mark of the partial exam (control)
    F = Mark of the final exam
    T = Mark of the practical work

    Exams mark = 0.6F+0.4P if F< P or F=P
    F if F>P

    Final mark = 40% T + 60% Exam mark

    Reassessment: only those people who, having taken the final exam (an NP is not valid), have an exam mark lower than 4 can present themselves for the reassessment. The maximum exam mark that can be obtained in the reassessment is 7.

    Bibliography

    Basic

    Previous capacities

    Knowledge of machine learning and basic AI algorithms.