Computational Tutorial: An introduction to LSTMs in Tensorflow
Description
Long Short Term Memory networks (LSTMs) are a type of recurrent neural network that can capture long term dependencies and are frequently used for natural language modeling and speech recognition. In this tutorial we will cover the conceptual basics of LSTMs and implement basic LSTM in tensorflow. The second part of the tutorial will give an introduction to the basics of tensorflow, an opensource software package used for implementing neural networks.
Instructions are here for installing python and tensorflow: https://github.com/nicholaslocascio/intro-deeplearning-setup
Please bring a laptop as there will be hands on exercises!
Please RSVP here so we know how much food to get! https://goo.gl/forms/jc9JmaSQ95lkgLK43
After the tutorial, slides and resources will be posted on the computational tutorials stellar page.
slides, references, and exercises: https://stellar.mit.edu/S/project/bcs-comp-tut/materials.html
videos: http://cbmm.mit.edu/videos?field_video_grouping_tid[0]=781
Other tutorial dates this spring:
5/10 2-4pm: Better Science Code taught by Eric Denovellis (46-3189)
Speaker Bio
Harini is an MEng student at CSAIL. Her current research uses LSTMs to model physiological timeseries.
Nick is MEng student at CSAIL working in Regina Barzilay's group. His current research is in applying deep learning to diagnostic mammography screening.