Deep Learning(CS7015): Lec 10.7 Hierarchical softmax

Показать описание

lec10mod07

NPTEL-NOC IITM
deep learning
prof.mitesh khapra
iitmadras
nptel deep learning

Рекомендации по теме

Комментарии

It really gets easier to understand when you see this (or any) lecture, the second time.

ashwiniked

If we have V nodes as leaf then we will be having V -1 nodes internally.

tusharrohilla

What does the vector for an internal node look like and what is the vector length for each internal node?

iidtxbc

What is the answer to his last question? "Do we need to consider anything while constructing the tree?" "Is there any problem in constructing it randomly?"

mr_easy

Why does this just cut off mid-sentence? The most interesting part is how the tree is constructed!

mattspeck

7:54 Why can we take the products of these probabilities? That require the events to be independent. Why are P(π(on)₁=1|vₛₐₜ) and P(π(on)₂=0|vₛₐₜ) independent?

mainakghosh

Deep Learning(CS7015): Lec 10.7 Hierarchical softmax

Deep Learning(CS7015): Lec 10.7 Hierarchical softmax

Learning Hierarchical Information Flow with Recurrent Neural Modules

Binay tree, Hierarchical softmax

Whiteboard Wednesdays - Training Different Networks Using Hierarchical CNN

Deep Learning(CS7015): Lec 10.6 Contrastive estimation

Introduction to Hierarchical Attention Networks

Deep Learning(CS7015): Lec 11.4 (Par-2) CNNs (success stories on ImageNet) (Contd.)

7 Hierarchical Softmax

Hierarchical multi-task learning for ctc based speech recognition

Deep Learning Part - II (CS7015): Lec 18.5 Unsupervised Learning with RBMs

Human-Computer QA: Attention-Based Convolutional Neural Network for Machine

Lecture 9 | (1/3) Convolutional Neural Networks

Ali Ghodsi, Deep Learning, Regularization, Fall 2023, Lecture 4,

Attention hierarchy in mindfulness practice

Pascal Vincent: Training neural networks in time independent of output layer size

Ali Ghodsi, Lec [5.1]: Deep Learning, Recurrent neural network

Mapping the output layer to Softmax

Processing Megapixel Images with Deep Attention-Sampling Models

Squeeze-and-Attention Networks for Semantic Segmentation

Christopher D Manning: A Neural Network Model That Can Reason (ICLR 2018 invited talk)

[Neural Network 3] Limitations of Perceptron and Multilayer Perceptron

David Schwab - How Noise Affects the Hessian Spectrum in Overparameterized Neural Networks

Discussion on the results obtained from word2vec

Log. Reg vs Perceptron | Lec 4 | Introduction to Neural Networks and Backpropagation | Deep Learning