Deep Learning Specialization (TensorFlow, Keras, PyTorch)

Understanding the core concepts of Deep Learning.

• What is Deep Learning? Machine Learning vs. Deep Learning.
• Artificial Neural Networks (ANNs).
• Activation Functions (ReLU, Sigmoid, Tanh).
• Gradient Descent and Backpropagation.

Building models with the Keras API.

• Introduction to TensorFlow 2.x.
• The Keras Sequential API.
• The Keras Functional API.
• Building and training your first neural network classifier.

Understanding the PyTorch framework.

• PyTorch Tensors.
• Autograd: Automatic differentiation.
• Building models with `nn.Module`.
• The training loop in PyTorch.

Improving model performance and preventing overfitting.

• Overfitting and Underfitting.
• Regularization (Dropout, L2).
• Optimizers (Adam, SGD, RMSprop).
• Hyperparameter tuning.

Understanding the building blocks of computer vision.

• The convolution operation.
• Padding and Stride.
• Pooling layers (Max, Average).
• Building a basic CNN for image classification (e.g., MNIST).

Using advanced CNNs for complex tasks.

• Classic architectures (VGG, ResNet).
• Transfer Learning with pre-trained models.
• Image augmentation techniques.

Handling time-series data and text.

• Handling sequence data.
• Recurrent Neural Networks (RNNs).
• The vanishing/exploding gradient problem.
• Text preprocessing (Tokenization, Embeddings).

Capturing long-term dependencies in data.

• Long Short-Term Memory (LSTM) networks.
• Gated Recurrent Units (GRU).
• Applications: Time-series forecasting.
• Applications: Text generation.

Understanding the "Attention Is All You Need" concept.

• The problem with RNNs/LSTMs.
• The concept of Self-Attention.
• The Transformer architecture (Encoder-Decoder).
• Overview of BERT and GPT models.

Building an end-to-end deep learning project.

• Project: Build an advanced image classifier or text generator.
• Choose either TensorFlow/Keras or PyTorch.
• Train, tune, and evaluate the model.
• Final project presentation and code review.