Using Recurrent Neural Networks and TensorFlow to Recognize Handwriting

Recurrent neural networks (RNNs) are designed to model sequential information and are widely used to solve the problems of speech recognition, language modeling, translation, and image captioning. At the recent TensorFlow Chicago meetup, it was discussed how to perform basic mathematical calculations or recognize handwriting using RNNs and TensorFlow together.

Teaching recurrent neural networks using TensorFlow

In his session, Rajiv Shah, Adjunct Assistant Professor at University of Illinois, provided a brief introduction to recurrent neural networks. Using TensorFlow’s code as an example, he also demonstrated how to:

• model a sine wave
• perform basic addition
• generate handwriting using

Rajiv has also answered some questions from the audience, sharing his ideas on what the prime reasons for choosing an RNN rather than a standard network are, what challenges one may face when training a model, etc.

Join the meetup group to stay informed about the upcoming events.

Want details? Watch the video!

Related slides

Further reading

• Recurrent Neural Networks: Classifying Diagnoses with Long Short-Term Memory
• Text Prediction with TensorFlow and Long Short-Term Memory—in Six Steps
• Building a Keras-Based Image Classifier Using TensorFlow for a Back End

About the speaker

Rajiv Shah is Data Scientist at a large insurance company and Adjunct Assistant Professor at the University of Illinois in Chicago. He is an active member of the data science community in Chicago with projects and publications related to surveillance and red light cameras. He has a PhD from the University of Illinois at Urbana Champaign.