RNNs for Time Series Data

Current Status
Not Enrolled
Get Started
This course is currently closed

Powerful predictive systems reliant on sequential data (various NLP tasks, time series analysis, signal processing) can be created using RNNs—or Recurrent Neural Networks. By the end of this course, students will learn the advanced methods of this complex topic, acquire practical skills to implement Recurrent Neural Networks, and build deep learning models using TensorFlow in order to solve real-world problems.


Data Society has designed this course for students who possess:

  • a strong foundation in Python

  • familiarity with common libraries: SciKit-Learn, Pandas, NumPy, and Matplotlib

  • comfort building machine learning models and foundational neural network programs


By the end of the course, students will be able to:

  • Understand which mathematical concepts make RNNs a suitable architecture for solving problems that are sequential in nature
  • Implement and train RNNs using TensorFlow
  • Optimize RNNs to improve overall performance


Hours of instruction: 2 hrs of video, 2 hrs of live learning

plus time for exercises