Deep Learning Architectures

This three-day workshop delves into advanced neural network architectures, including Residual Networks (ResNets), DenseNets, EfficientNet, NasNet, LSTM, GRUs, and Transformers. Participants will engage in hands-on sessions and case studies to understand the practical applications of these architectures.

To provide PhD scholars and academicians with advanced skills in designing and optimizing deep learning architectures. This course aims to enhance understanding of advanced neural networks, CNNs, and RNNs, focusing on their applications and optimization techniques.

Day 1: Advanced Neural Networks

• Lecture Topics:
  • Architectures beyond feedforward networks: Residual Networks (ResNets), DenseNets
  • Optimization techniques and activation functions
• Discussion & Case Studies:
  • Research insights on optimizing deep networks
  • Interactive session on designing custom architectures

Day 2: Advanced Convolutional Neural Networks (CNNs)

• Lecture Topics:
  • Latest advancements in CNNs: EfficientNet, NasNet
  • Transfer learning and fine-tuning pre-trained models
• Discussion & Case Studies:
  • Applications in medical imaging and autonomous vehicles
  • Practical implementation and best practices

Day 3: Advanced Recurrent Neural Networks (RNNs)

• Lecture Topics:
  • Beyond basic RNNs: Long Short-Term Memory (LSTM), Gated Recurrent Units (GRUs)
  • Attention mechanisms and Transformers
• Discussion & Case Studies:
  • Case studies in natural language generation and time-series forecasting
  • Hands-on session with sequential data

• Design and optimize advanced neural network architectures.
• Implement and fine-tune state-of-the-art CNN models.
• Develop advanced RNN models with attention mechanisms.
• Apply deep learning techniques to practical applications in various fields.
• Conduct high-level research in deep learning and AI.

Data scientists, machine learning engineers, AI researchers, and academicians in AI and machine learning.