An interpretation of the final fully connected layer

05/24/2022
by   Siddhartha, et al.
180

In recent years neural networks have achieved state-of-the-art accuracy for various tasks but the the interpretation of the generated outputs still remains difficult. In this work we attempt to provide a method to understand the learnt weights in the final fully connected layer in image classification models. We motivate our method by drawing a connection between the policy gradient objective in RL and supervised learning objective. We suggest that the commonly used cross entropy based supervised learning objective can be regarded as a special case of the policy gradient objective. Using this insight we propose a method to find the most discriminative and confusing parts of an image. Our method does not make any prior assumption about neural network achitecture and has low computational cost. We apply our method on publicly available pre-trained models and report the generated results.

READ FULL TEXT

page 1

page 2

page 3

page 4

03/16/2021

Hebbian Semi-Supervised Learning in a Sample Efficiency Setting

We propose to address the issue of sample efficiency, in Deep Convolutio...
04/12/2019

Similarities between policy gradient methods (PGM) in Reinforcement learning (RL) and supervised learning (SL)

Reinforcement learning (RL) is about sequential decision making and is t...
10/05/2021

Quasi-Newton policy gradient algorithms

Policy gradient algorithms have been widely applied to reinforcement lea...
11/19/2018

An Efficient Transfer Learning Technique by Using Final Fully-Connected Layer Output Features of Deep Networks

In this paper, we propose a computationally efficient transfer learning ...
09/20/2022

CoV-TI-Net: Transferred Initialization with Modified End Layer for COVID-19 Diagnosis

This paper proposes transferred initialization with modified fully conne...
06/25/2021

Efficient Document Image Classification Using Region-Based Graph Neural Network

Document image classification remains a popular research area because it...

Code Repositories

interpret-fc-layer

An interpretation of the final fully connected layer in pre-trained imagenet models


view repo

Please sign up or login with your details

Forgot password? Click here to reset