Dive into advanced natural language processing concepts in this lecture from Carnegie Mellon University's CS 11-711 course. Explore neural networks and their toolkits, learn to define differentiable functions, and master the forward and backward algorithms. Gain insights into parameter updates and practical training techniques. Enhance your understanding of text classification through in-depth discussions on topics like bag of words, tensor data structures, initialization methods, and strategies to combat overfitting.
Overview
Syllabus
Intro
Bag of Words
Algorithm Sketch
Tensors
Tensor Data Structure
initialization
uniform initialization
other methods
model
computation
operations
rectified linear unit
chain rule
back propagation
backward code
parameter updates
update sparse
edigrad
atom
training tricks
overfitting
Taught by
Graham Neubig
Related Courses
-
Neural Nets for NLP 2021 - Building A Neural Network Toolkit for NLP
-
Neural Nets for NLP - Class Introduction & Why Neural Nets?
-
Neural Nets for NLP 2021 - Language Modeling, Efficiency/Training Tricks
-
CMU Neural Nets for NLP 2018 - Conditioned Generation
-
Neural Nets for NLP 2017 - A Simple Exercise - Predicting the Next Word in a Sentence
-
Neural Nets for NLP 2020 - Language Modeling, Efficiency/Training Tricks
