7,862 Hits in 3.6 sec

Improved Machine Learning using Adaptive Boosting algorithm in Membrane Protein Prediction

Decision making for predicting the identification of membrane protein types was performed using an algorithm framework to improve the learning accuracy, by putting the training samples weights in the learning  ...  The performance of different ensemble classifiers such as Random Forest, AdaBoost, is analyzed.  ...  The transfer learning model is one of another important key factor for learning success.  ... 
doi:10.35940/ijitee.k2207.1081219 fatcat:xxmjmcfp4femvg3b2fh3dhdzey

Hypothesis Disparity Regularized Mutual Information Maximization [article]

Qicheng Lao, Xiang Jiang, Mohammad Havaei
2020 arXiv   pre-print
HDMI achieves state-of-the-art adaptation performance on benchmark datasets for UDA in the context of HTL, without the need to access the source data during the adaptation.  ...  We propose a hypothesis disparity regularized mutual information maximization~(HDMI) approach to tackle unsupervised hypothesis transfer -- as an effort towards unifying hypothesis transfer learning (HTL  ...  Relation to unsupervised domain adaptation Unsupervised domain adaptation, also considered as a form of transfer learning (transductive transfer learning (Pan and Yang 2009 )), aims to adapt a target  ... 
arXiv:2012.08072v1 fatcat:iyofolrwajffrjwkxtnxzn2eau

Multiple Expert Brainstorming for Domain Adaptive Person Re-identification [article]

Yunpeng Zhai, Qixiang Ye, Shijian Lu, Mengxi Jia, Rongrong Ji, Yonghong Tian
2020 arXiv   pre-print
Often the best performing deep neural models are ensembles of multiple base-level networks, nevertheless, ensemble learning with respect to domain adaptive person re-ID remains unexplored.  ...  In this paper, we propose a multiple expert brainstorming network (MEB-Net) for domain adaptive person re-ID, opening up a promising direction about model ensemble problem under unsupervised conditions  ...  Acknowledgement This work is partially supported by grants from the National Key R&D Program of China under grant 2017YFB1002400, the National Natural Science Foundation of China (NSFC) under contract  ... 
arXiv:2007.01546v3 fatcat:m66wzjbigjbxtohccg4psa4oue

Cross-Lingual Adaptation for Type Inference [article]

Zhiming Li, Xiaofei Xie, Haoliang Li, Zhengzi Xu, Yi Li, Yang Liu
2021 arXiv   pre-print
In this paper, we propose cross-lingual adaptation of program analysis, which allows us to leverage prior knowledge learned from the labeled dataset of one language and transfer it to the others.  ...  Besides, by leveraging data from strongly typed languages, PLATO improves the perplexity of the backbone cross-programming-language model and the performance of downstream cross-lingual transfer for type  ...  Transfer learning and domain adaptation is becoming increasingly popular, where a model developed for a task (or domain) is reused as the starting point for training a model for another task (or domain  ... 
arXiv:2107.00157v1 fatcat:elq3ytr7g5glxlypk7gkeapcny

A2J: Anchor-to-Joint Regression Network for 3D Articulated Pose Estimation from a Single Depth Image [article]

Fu Xiong, Boshen Zhang, Yang Xiao, Zhiguo Cao, Taidong Yu, Joey Tianyi Zhou, Junsong Yuan
2019 arXiv   pre-print
For 3D hand and body pose estimation task in depth image, a novel anchor-based approach termed Anchor-to-Joint regression network (A2J) with the end-to-end learning ability is proposed.  ...  They contribute to predict the positions of the joints in ensemble way to enhance generalization ability.  ...  We also thank the anonymous reviewers for their suggestions to enhance the quality of this paper.  ... 
arXiv:1908.09999v1 fatcat:ob3ztyrkqrdyjij7r532lvquk4

Unsupervised and self-adaptative techniques for cross-domain person re-identification [article]

Gabriel Bertocco and Fernanda Andaló and Anderson Rocha
2022 arXiv   pre-print
For evaluation, we consider three well-known deep learning architectures and combine them for final decision-making.  ...  We also introduce a new self-ensembling strategy, in which weights from different iterations are aggregated to create a final model combining knowledge from distinct moments of the adaptation.  ...  ACKNOWLEDGMENT We thank the financial support of the São Paulo Research Foundation (FAPESP) through the grants DéjàVu #2017/12646-3 and #2019/15825-1.  ... 
arXiv:2103.11520v3 fatcat:hzwleq77wfamxca7tq5o5ntkju

Deep Class Incremental Learning from Decentralized Data [article]

Xiaohan Zhang, Songlin Dong, Jinjie Chen, Qi Tian, Yihong Gong, Xiaopeng Hong
2022 arXiv   pre-print
Secondly, we introduce a paradigm to create a basic decentralized counterpart of typical (centralized) class-incremental learning approaches, and as a result, establish a benchmark for the DCIL study.  ...  In this paper, we focus on a new and challenging decentralized machine learning paradigm in which there are continuous inflows of data to be addressed and the data are stored in multiple repositories.  ...  DNNs in the training steps to avoid co-adaptations of these nodes.  ... 
arXiv:2203.05984v1 fatcat:vkxqlew7ivef3cocl7rb5mtasm

A Review of Artificial Intelligence Technologies for Early Prediction of Alzheimer's Disease [article]

Kuo Yang, Emad A. Mohammed
2020 arXiv   pre-print
, Graph CNN (GCN), Ensemble Learning, and Transfer Learning.  ...  outlook for this research.  ...  Ensemble Learning for AD Prediction F. Li et al. applied the ensemble learning method to predict the early stage of AD [4] .  ... 
arXiv:2101.01781v1 fatcat:lqtovw4jlbdcjhh5rvresp2gmq

Learning Test-time Augmentation for Content-based Image Retrieval [article]

Osman Tursun, Simon Denman, Sridha Sridharan, Clinton Fookes
2021 arXiv   pre-print
effective for improving performance, and is practical, and transferable.  ...  Experimental results on large trademark retrieval (METU trademark dataset) and landmark retrieval (ROxford5k and RParis6k scene datasets) tasks show that the learned ensemble of transformations is highly  ...  to learn the ensemble of TTA.  ... 
arXiv:2002.01642v4 fatcat:tz54urgk4rf3djfzt6uh2lerhe

A Comprehensive Survey on Transfer Learning [article]

Fuzhen Zhuang, Zhiyuan Qi, Keyu Duan, Dongbo Xi, Yongchun Zhu, Hengshu Zhu, Hui Xiong, Qing He
2020 arXiv   pre-print
In order to show the performance of different transfer learning models, over twenty representative transfer learning models are used for experiments.  ...  And the experimental results demonstrate the importance of selecting appropriate transfer learning models for different applications in practice.  ...  For example, Ensemble Framework of Anchor Adapters (ENCHOR) [133] is a weighting ensemble framework proposed by Zhuang et al. An anchor is a specific instance.  ... 
arXiv:1911.02685v3 fatcat:oeofarz7tnbtlblvta4evx3e34

DiVA: Diverse Visual Feature Aggregation for Deep Metric Learning [article]

Timo Milbich, Karsten Roth, Homanga Bharadhwaj, Samarth Sinha, Yoshua Bengio, Björn Ommer, Joseph Paul Cohen
2020 arXiv   pre-print
For effective generalization, however, such an image representation needs to capture a diverse range of data characteristics.  ...  Deep metric learning (DML) is a powerful framework for learning such similarities which not only generalize from training data to identically distributed test distributions, but in particular also translate  ...  For the latter, we adapt contrastive self-supervised learning to the needs of supervised DML.  ... 
arXiv:2004.13458v4 fatcat:flyvghpkgraivfun5szrcraw7u

RetinaGAN: An Object-aware Approach to Sim-to-Real Transfer [article]

Daniel Ho and Kanishka Rao and Zhuo Xu and Eric Jang and Mohi Khansari and Yunfei Bai
2021 arXiv   pre-print
The success of deep reinforcement learning (RL) and imitation learning (IL) in vision-based robotic manipulation typically hinges on the expense of large scale data collection.  ...  RetinaGAN improves upon the performance of prior sim-to-real methods for RL-based object instance grasping and continues to be effective even in the limited data regime.  ...  We thank Chris Harris and Alex Irpan for comments on the manuscript.  ... 
arXiv:2011.03148v2 fatcat:a2c7ni6ln5ez5nd3zgbfpxhiia

FocalMix: Semi-Supervised Learning for 3D Medical Image Detection [article]

Dong Wang, Yuan Zhang, Kexin Zhang, Liwei Wang
2020 arXiv   pre-print
In this paper, we propose a novel method, called FocalMix, which, to the best of our knowledge, is the first to leverage recent advances in semi-supervised learning (SSL) for 3D medical image detection  ...  Results show that our proposed SSL methods can achieve a substantial improvement of up to 17.3% over state-of-the-art supervised learning approaches with 400 unlabeled CT scans.  ...  Then, we aggregate these guessed targets for every anchor by the average ensemble.  ... 
arXiv:2003.09108v1 fatcat:jdb2wrha2bfxnpexbmlirswkxq

A Fortran-Keras Deep Learning Bridge for Scientific Computing [article]

Jordan Ott, Mike Pritchard, Natalie Best, Erik Linstead, Milan Curcic, Pierre Baldi
2020 arXiv   pre-print
In this context, FKB enables a hyperparameter search of one hundred plus candidate models of subgrid cloud and radiation physics, initially implemented in Keras, to be transferred and used in Fortran.  ...  These software libraries come pre-loaded with a variety of network architectures, provide autodifferentiation, and support GPUs for fast and efficient computation.  ...  Having described the deep learning anchor within Python, the next section develops the foundation for anchoring the bridge with Fortran.  ... 
arXiv:2004.10652v2 fatcat:7esikvrubfauneeavgbecrqdsi

Polyp Detection and Segmentation using Mask R-CNN: Does a Deeper Feature Extractor CNN Always Perform Better?

Hemin Ali Qadir, Younghak Shin, Johannes Solhusvik, Jacob Bergsland, Lars Aabakken, Ilangko Balasingham
2019 2019 13th International Symposium on Medical Information and Communication Technology (ISMICT)  
Finally, we propose an ensemble method for further performance improvement. We evaluate the performance on the 2015 MICCAI polyp detection dataset.  ...  In this paper, we adapt Mask R-CNN and evaluate its performance with different modern convolutional neural networks (CNN) as its feature extractor for polyp detection and segmentation.  ...  Therefore, we use transfer learning by initializing the weights of our CNN feature extractors from models pre-trained on Microsoft's COCO dataset [28] .  ... 
doi:10.1109/ismict.2019.8743694 dblp:conf/ismict/QadirSSBAB19 fatcat:2nazmpa3cvaclgosubnb22yn6i
« Previous Showing results 1 — 15 out of 7,862 results