Interpretable Embedding Procedure Knowledge Transfer via Stacked Principal Component Analysis and Graph Neural Network

04/28/2021
by   Seunghyun Lee, et al.
16

Knowledge distillation (KD) is one of the most useful techniques for light-weight neural networks. Although neural networks have a clear purpose of embedding datasets into the low-dimensional space, the existing knowledge was quite far from this purpose and provided only limited information. We argue that good knowledge should be able to interpret the embedding procedure. This paper proposes a method of generating interpretable embedding procedure (IEP) knowledge based on principal component analysis, and distilling it based on a message passing neural network. Experimental results show that the student network trained by the proposed KD method improves 2.28 dataset, which is higher performance than the state-of-the-art (SOTA) method. We also demonstrate that the embedding procedure knowledge is interpretable via visualization of the proposed KD process. The implemented code is available at https://github.com/sseung0703/IEPKT.

READ FULL TEXT
research
07/04/2019

Graph-based Knowledge Distillation by Multi-head Attention Network

Knowledge distillation (KD) is a technique to derive optimal performance...
research
06/11/2020

Interpretable Visualizations with Differentiating Embedding Networks

We present a visualization algorithm based on a novel unsupervised Siame...
research
06/27/2021

Interpretable Network Representation Learning with Principal Component Analysis

We consider the problem of interpretable network representation learning...
research
07/04/2019

Graph-based Knowledge Distillation by Multi-head Self-attention Network

Knowledge distillation (KD) is a technique to derive optimal performance...
research
11/06/2017

Interpretable Feature Recommendation for Signal Analytics

This paper presents an automated approach for interpretable feature reco...
research
07/22/2018

On Tree-structured Multi-stage Principal Component Analysis (TMPCA) for Text Classification

A novel sequence-to-vector (seq2vec) embedding method, called the tree-s...
research
08/12/2021

Distilling Holistic Knowledge with Graph Neural Networks

Knowledge Distillation (KD) aims at transferring knowledge from a larger...

Please sign up or login with your details

Forgot password? Click here to reset