Introduction to Deep Learning and Generative Artificial Intelligence
Short Courses
Enrollment and Registration: 100% free course
Tuition Fees: No costs
General Objectives
Present participants with the fundamentals and applications of AI models based on deep learning, with emphasis on Convolutional Neural Networks (CNNs), LSTM-type Recurrent Networks, and Transformers, which form the basis of advanced models like ChatGPT.
The course aims to provide a practical and theoretical understanding of these tools, allowing participants to develop the ability to apply these types of models in various domains.
Specific Objectives
Understand the fundamental concepts of Deep Learning, focusing on CNNs, LSTMs, and Transformers.
Apply Deep Learning techniques to real-world problems, such as image recognition, time-series analysis, and text generation/transformation.
Effectively use pre-trained models to achieve accurate and efficient results, taking advantage of models available online.
Build and refine their own Deep Learning models according to the specific requirements of different tasks.
Interpret the results of complex models and use this information for informed decision-making.
Understand the capabilities and limitations of generative models, both in language and in generating other content, such as images and voice.
Target Audience
Professionals and students in the areas of Information Technology, Data Science, Engineering, Mathematics, and related fields who wish to deepen their knowledge of Deep Learning and Generative Artificial Intelligence.
Artificial intelligence enthusiasts who want to understand the practical applications of technologies such as CNNs, LSTMs, and Transformers.
People looking for a practical and theoretical introduction to Deep Learning techniques, even without prior experience in the field.
Anyone interested in understanding how tools like ChatGPT, image generation models, and time-series forecasting are developed and applied in different contexts.
Program Content
- Computer Vision with CNNs (1st Session / 3 hours)
1.1 Introduction to Deep Learning and Convolutional Neural Networks (CNNs).
1.2 Practical examples:
1.2.1 Clothing item recognition.
1.2.2 Skin signal classification.
- Time Series with LSTM (2nd Session / 3 hours)
2.1 LSTM-type Recurrent Neural Networks for pattern and trend analysis.
2.2 Practical example:
2.2.1 S&P 500 Index Price Forecast.
- Deep Learning for Text with Transformers (3rd Session / 3 hours)
3.1 Functioning of models like ChatGPT and their application in natural language processing.
3.2 Practical examples:
3.2.1 Experimentation with OpenAI Playground.
3.2.2 Exploration of the Hugging Face library.
- Other Generative AI Examples (4th Session / 3 hours)
4.1 Image generation using advanced models like DALL-E 3 and Stable Diffusion XL.
4.2 Tools:
4.2.1 Libraries like diffusers and KerasCV for creating visual content.
Training Dates/Times
4 sessions of 3 hours, in the evening
(Date and time to be confirmed shortly)
Total Training Duration
12 hours
Methodology
The course "Introduction to Deep Learning and Generative Artificial Intelligence" adopts a practical and applied approach, combining theoretical exposure with practical examples to maximize participants' understanding and experience.
During each session, participants will have access to content that includes step-by-step demonstrations and practical exercises, which can be followed either in person or remotely via Zoom. The use of cloud-based learning tools will allow participants to test models and techniques without the need for high-performance equipment, ensuring an accessible learning experience.
Additionally, participants will have access to support materials, such as theoretical summaries, demonstrations, online playgrounds, and code repositories on GitHub, for reference throughout the course and for independent study during and after each session.
Instructor
Luís Cunha, PhD
Professor of the Computer Science Bachelor's Degree at ISMT
Required Documentation
Location: In-person at ISMT or online via Zoom (link sent by email).
Requirements: Computer or tablet to follow the examples.
Slots: 30 participants.
Certificate: Issued to participants who complete all sessions.
Organization: Scientific Coordination of the Computer Science Bachelor's Degree at ISMT.
For more information:
Phone: (+351) 239 488 030
Email: cursosinformatica@ismt.pt
Instituto Superior Miguel Torga
Largo da Cruz de Celas, 1
3000-132 Coimbra