Neural Networks, Transfer Learning, and Visualizing Graphs with TensorFlow

by Sophia TurolApril 14, 2016
Learn what's behind the architecture of a convolutional neural network, how it can be implemented with TensorFlow, how to visualize graphs with TensorBoard, etc.


Below are the videos from the Boston TensorFlow meetup—sponsored and organized by Altoros on March 7, 2016.


Talk #1: Neural Networks with Google TensorFlow

In his session, Darshan Patel embraced computer vision tasks, convolution neural network’s (CNN) architecture, explaining how to implement it with TensorFlow, and TensorBoard graph visualization.



Talk #2: Transfer learning on TensorFlow in 30 minutes

Syed Tousif Ahmed of Alhold USA provided answers for the following questions:

  • What is transfer learning (strategies, use cases)?
  • How to collect data?
  • How to retrain a model in TensorFlow?



Fireside chat

Later on, Renat Khasanshyn interviewed Darshan Patel and Syed Tousif Ahmed on the following aspects:

  • What does attract you most in TensorFlow?
  • How can TensorFlow make a difference as a library in the next 5–10 years?
  • Pieces of advice for those getting started with TensorFlow



Join our group to get notified about the upcoming events!


Further reading


About the experts

Darshan Patel is a machine learning enthusiast. He is in his second year of Computer Science Masters in the College of Computer and Information science at Northeastern University. Prior to a master study, Darshan was a software developer at TCS, India.

Syed Tousif Ahmed is a Computer Vision Developer working at Ahold USA. He is also a student in Computer Engineering at Rochester Institute of Technology. Syed’s area of interests are computer vision, machine learning, cryptography, and Internet of Things.