Learning Deep Learning: Unit 2

Overview

This course covers advanced deep learning topics, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and modern language models. You will learn techniques for image classification, time series prediction, and natural language processing. The course includes building and optimizing CNNs for image recognition, using architectures such as AlexNet, VGGNet, GoogLeNet, and ResNet, and working with pre-trained models. You will also work with RNNs and LSTMs for tasks like forecasting and text autocompletion. The curriculum covers neural language models, word embeddings (such as Word2vec and wordpieces), encoder-decoder architectures, attention mechanisms, and Transformers for machine translation. Hands-on projects using TensorFlow and PyTorch will help you develop practical skills for solving real-world problems in computer vision and language processing.

Syllabus

Learning Deep Learning: Unit 2

This module provides a comprehensive introduction to advanced deep learning techniques for processing images and natural language. It covers convolutional neural networks for image classification, including architectures like AlexNet, VGGNet, GoogLeNet, and ResNet. The module then explores recurrent neural networks and LSTMs for time series and sequential data, followed by neural language models and word embeddings. Finally, it introduces encoder-decoder architectures, attention mechanisms, and Transformer models for neural machine translation, with practical implementations in TensorFlow and PyTorch throughout.