loading
In-training and post-training generalization methods: The case of pparα and pparγ agonists
2015 Edition, July 1, 2015 - IEEE - Institute of Electrical and Electronics Engineers, Inc.

In this paper, the effects of regularization on the generalization capabilities of a neural network model are analyzed. We compare the performance of Levenberg-Marquardt and Bayesian Regularization algorithms with and without...

Understanding Mixup Training Methods
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Mixup is a neural network training method that generates new samples by linear interpolation of multiple samples and their labels. The mixup training method has better generalization ability than the traditional Empirical Risk...

Improvement of Generalization Ability of Deep CNN via Implicit Regularization in Two-stage Training Process
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Optimization of deep learning is no longer an imminent problem, due to various gradient descent methods and the improvements of network structure, including activation functions, the connectivity style, etc. Then the actual application depends on...

Graph Adversarial Training: Dynamically Regularizing Based on Graph Structure
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Recent efforts show that neural networks are vulnerable to small but intentional perturbations on input features in visual classification tasks. Due to the additional consideration of connections between examples (e.g., articles with citation link tend to be in...

Training Data Selection and Update Strategies for Airborne Post-Doppler STAP
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Space-time adaptive processing (STAP) of multichannel radar data is an established and powerful method for detecting ground moving targets, as well as for estimating their geographical positions and line-of-sight velocities. Crucial steps for practical...

Normalization in Training U-Net for 2D Biomedical Semantic Segmentation
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

2D biomedical semantic segmentation is important for robotic vision in surgery. Segmentation methods based on Deep Convolutional Neural Network (DCNN) can out-perform conventional methods in terms of both accuracy and levels of automation. One common...

Robot-assisted Training in Laparoscopy using Deep Reinforcement Learning
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Minimally Invasive Surgery (MIS) is increasingly becoming a vital method of reducing surgical trauma and significantly improving post-operative recovery. However, skillful handling of surgical instruments used in MIS, especially for laparoscopy, requires a...

Impact of Using a Robot Patient for Nursing Skill Training in Patient Transfer
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

In the past few decades, simulation training has been used to help nurses improve their patient-transfer skills. However, the effectiveness of such training remains limited because it lacks effective ways of simulating patients' actions realistically...

Development of a Novel Home Based Multi-scene Upper Limb Rehabilitation Training and Evaluation System for Post-stroke Patients
Volume PP - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Upper limb rehabilitation requires long-term, repetitive rehabilitation training and assessment, and many patients cannot pay for expensive medical fees in the hospital for so long time. It is necessary to design an effective, low cost, and reasonable home...

Improvement of Generalization Ability of Deep CNN via Implicit Regularization in Two-Stage Training Process
2018 Edition, Volume 6, January 1, 2018 - IEEE - Institute of Electrical and Electronics Engineers, Inc.

Optimization of deep learning is no longer an imminent problem, due to various gradient descent methods and the improvements of network structure, including activation functions, the connectivity style, and so on. Then the actual application...

Advertisement