Updated in May 2025.

This course now features Coursera Coach! A smarter way to learn with interactive, real-time conversations that help you test your knowledge, challenge assumptions, and deepen your understanding as you progress through the course. Recurrent Neural Networks (RNNs) are a powerful class of neural networks designed for sequence data, making them ideal for time series prediction and natural language processing tasks. This course begins with an introduction to the fundamental concepts of RNNs and explores their application in forecasting and time series prediction. You will delve into coding with TensorFlow, learning how to implement autoregressive models and simple RNNs for various predictive tasks. As the course progresses, you will encounter more sophisticated RNN architectures such as GRUs and LSTMs. These units are essential for handling complex sequences and long-distance dependencies in data. Practical sessions will guide you through using these models for challenging tasks, including stock return prediction and image classification on the MNIST dataset. The course also covers the critical aspect of managing data shapes and ensuring your models are well-structured and efficient. Towards the end, the course shifts focus to natural language processing (NLP), where you will explore embeddings, text preprocessing, and text classification using LSTMs. By combining theoretical knowledge with hands-on coding exercises, you will develop a robust understanding of how to leverage RNNs for various applications. Whether you are predicting stock prices or classifying text, this course equips you with the skills needed to succeed in the field of deep learning. This course is ideal for data scientists, machine learning engineers, and AI enthusiasts who want to learn and implement recurrent neural networks for time series analysis and natural language processing. Basic knowledge of Python and TensorFlow is recommended.