Explicit inductive bias for transfer learning with convolutional networks
LI Xuhong, Y Grandvalet… - … Conference on Machine …, 2018 - proceedings.mlr.press
In inductive transfer learning, fine-tuning pre-trained convolutional networks substantially
outperforms training from scratch. When using fine-tuning, the underlying assumption is that …
outperforms training from scratch. When using fine-tuning, the underlying assumption is that …
Deep convolutional neural networks for massive MIMO fingerprint-based positioning
This paper provides an initial investigation on the application of convolutional neural networks
(CNNs) for fingerprint-based positioning using measured massive MIMO channels. When …
(CNNs) for fingerprint-based positioning using measured massive MIMO channels. When …
Interpretable deep learning: Interpretation, interpretability, trustworthiness, and beyond
Deep neural networks have been well-known for their superb handling of various machine
learning and artificial intelligence tasks. However, due to their over-parameterized black-box …
learning and artificial intelligence tasks. However, due to their over-parameterized black-box …
Formulation design, challenges, and development considerations for fixed dose combination (FDC) of oral solid dosage forms
D Desai, J Wang, H Wen, X Li… - Pharmaceutical …, 2013 - Taylor & Francis
Fixed dose combination (FDC) products are common in the treatment of hypertension, diabetes,
human immunodeficiency virus, and tuberculosis. They make it possible to combine two …
human immunodeficiency virus, and tuberculosis. They make it possible to combine two …
[HTML][HTML] Drug delivery approaches in addressing clinical pharmacology-related issues: opportunities and challenges
H Wen, H Jung, X Li - The AAPS journal, 2015 - Springer
Various drug delivery approaches can be used to maximize therapeutic efficacy and minimize
side effects, by impacting absorption, distribution, metabolism, and elimination (ADME) of …
side effects, by impacting absorption, distribution, metabolism, and elimination (ADME) of …
From distributed machine learning to federated learning: A survey
In recent years, data and computing resources are typically distributed in the devices of end
users, various regions or organizations. Because of laws or regulations, the distributed data …
users, various regions or organizations. Because of laws or regulations, the distributed data …
[HTML][HTML] Electrically tunable two-dimensional heterojunctions for miniaturized near-infrared spectrometers
Miniaturized spectrometers are of considerable interest for their portability. Most designs to
date employ a photodetector array with distinct spectral responses or require elaborated …
date employ a photodetector array with distinct spectral responses or require elaborated …
Dual optimization of bulk and surface via guanidine halide for efficient and stable 2D/3D hybrid perovskite solar cells
In order to improve both performance and stability of perovskite solar cells, a design is
provided by combining the advantages of high‐efficiency 3D perovskite solar cells (PSCs) and …
provided by combining the advantages of high‐efficiency 3D perovskite solar cells (PSCs) and …
[PDF][PDF] Association of estrogen receptor α polymorphisms with susceptibility to chronic hepatitis B virus infection
G Deng, G Zhou, Y Zhai, S Li, X Li, Y Li, R Zhang… - …, 2004 - Wiley Online Library
Several studies have demonstrated that estrogen receptor α (ESR1) participates in the
pathogenesis of persistent hepatitis B virus (HBV) infection. To examine whether polymorphisms …
pathogenesis of persistent hepatitis B virus (HBV) infection. To examine whether polymorphisms …
Transfer learning in computer vision tasks: Remember where you come from
Fine-tuning pre-trained deep networks is a practical way of benefiting from the representation
learned on a large database while having relatively few examples to train a model. This …
learned on a large database while having relatively few examples to train a model. This …