Filters








10 Hits in 1.6 sec

Diversity regularization in deep ensembles [article]

Changjian Shui, Azadeh Sadat Mozafari, Jonathan Marek, Ihsen Hedhli, Christian Gagné
<span title="2018-02-22">2018</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Calibrating the confidence of supervised learning models is important for a variety of contexts where the certainty over predictions should be reliable. However, it has been reported that deep neural network models are often too poorly calibrated for achieving complex tasks requiring reliable uncertainty estimates in their prediction. In this work, we are proposing a strategy for training deep ensembles with a diversity function regularization, which improves the calibration property while maintaining a similar prediction accuracy.
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1802.07881v1">arXiv:1802.07881v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/evbca7ruu5d7daiqqo44dojufa">fatcat:evbca7ruu5d7daiqqo44dojufa</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20191025052939/https://arxiv.org/pdf/1802.07881v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/3f/9b/3f9b2da32b9597978990f033d90e3cc37a14c316.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1802.07881v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

A Novel Unsupervised Post-Processing Calibration Method for DNNS with Robustness to Domain Shift [article]

Azadeh Sadat Mozafari, Hugo Siqueira Gomes, Christian Gagne
<span title="2019-11-25">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
The uncertainty estimation is critical in real-world decision making applications, especially when distributional shift between the training and test data are prevalent. Many calibration methods in the literature have been proposed to improve the predictive uncertainty of DNNs which are generally not well-calibrated. However, none of them is specifically designed to work properly under domain shift condition. In this paper, we propose Unsupervised Temperature Scaling (UTS) as a robust
more &raquo; ... n method to domain shift. It exploits unlabeled test samples instead of the training one to adjust the uncertainty prediction of deep models towards the test distribution. UTS utilizes a novel loss function, weighted NLL, which allows unsupervised calibration. We evaluate UTS on a wide range of model-datasets to show the possibility of calibration without labels and demonstrate the robustness of UTS compared to other methods (e.g., TS, MC-dropout, SVI, ensembles) in shifted domains.
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1911.11195v1">arXiv:1911.11195v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/rdyqsfr4ubebhoprttlfkehumy">fatcat:rdyqsfr4ubebhoprttlfkehumy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200929031745/https://arxiv.org/pdf/1911.11195v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/f8/91/f891d0e41857e8b214616e199acdc906349091d9.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1911.11195v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Controlling Over-generalization and its Effect on Adversarial Examples Generation and Detection [article]

Mahdieh Abbasi, Arezoo Rajabi, Azadeh Sadat Mozafari, Rakesh B. Bobba, Christian Gagne
<span title="2018-10-03">2018</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Convolutional Neural Networks (CNNs) significantly improve the state-of-the-art for many applications, especially in computer vision. However, CNNs still suffer from a tendency to confidently classify out-distribution samples from unknown classes into pre-defined known classes. Further, they are also vulnerable to adversarial examples. We are relating these two issues through the tendency of CNNs to over-generalize for areas of the input space not covered well by the training set. We show that
more &raquo; ... CNN augmented with an extra output class can act as a simple yet effective end-to-end model for controlling over-generalization. As an appropriate training set for the extra class, we introduce two resources that are computationally efficient to obtain: a representative natural out-distribution set and interpolated in-distribution samples. To help select a representative natural out-distribution set among available ones, we propose a simple measurement to assess an out-distribution set's fitness. We also demonstrate that training such an augmented CNN with representative out-distribution natural datasets and some interpolated samples allows it to better handle a wide range of unseen out-distribution samples and black-box adversarial examples without training it on any adversaries. Finally, we show that generation of white-box adversarial attacks using our proposed augmented CNN can become harder, as the attack algorithms have to get around the rejection regions when generating actual adversaries.
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1808.08282v2">arXiv:1808.08282v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/jks2ron3cfcjbavht3f5wprkgu">fatcat:jks2ron3cfcjbavht3f5wprkgu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200825003529/https://arxiv.org/pdf/1808.08282v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/6c/4f/6c4f5c95004cdfde3b58d543b8cb25d6b91ebed6.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1808.08282v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

A Framework for Knowledge Management System in the Cloud Computing Environment and Web 2.0

Seyed Hossein Siadat, Azadeh Sadat Mozafari Mehr
<span title="">2018</span> <i title="Iranian Research Institute for Information and Technology"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/cge2muindbckditjezwyatkb6m" style="color: black;">Iranian Journal of Information Processing &amp; Management</a> </i> &nbsp;
Today, data, information and knowledge are very important assets for organizations and the effective management of knowledge is considered a way to gain and sustain a competitive advantage in a highly dynamic environment of the organizations. With the growth of information and communication technologies, cloud computing and Web 2.0, as new phenomena, recommend helpful solutions in the field of knowledge management. In this article, we introduce and review various models of the cloud-based
more &raquo; ... dge management systems at first, and then we present a framework for knowledge management system in the cloud-computing environment. The proposed framework consists of seven main components. To compare to the other reviewed models, in this framework, additionally, we used the numerous benefits of cloud computing such as reducing the cost of hardware and software, flexibility, accessibility at any time and place, collaboration, etc. where we tried to concerning the great capability of cloud computing for gathering knowledge and providing business intelligence infrastructure.
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://doaj.org/article/b7477f2e055f4be0b0310520e444fe20">doaj:b7477f2e055f4be0b0310520e444fe20</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/alauot5wxrbqlj3xuj67fggsta">fatcat:alauot5wxrbqlj3xuj67fggsta</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20180421012152/http://jipm.irandoc.ac.ir/files/site1/user_files_e1671e/azadehmozafarimehr-A-10-3747-1-b988e7a.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ff/71/ff71bb92e88557811bf78eb3d27e70c7b3d42e86.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a>

Attended Temperature Scaling: A Practical Approach for Calibrating Deep Neural Networks [article]

Azadeh Sadat Mozafari, Hugo Siqueira Gomes, Wilson Leão, Steeven Janny, Christian Gagné
<span title="2019-05-08">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Recently, Deep Neural Networks (DNNs) have been achieving impressive results on wide range of tasks. However, they suffer from being well-calibrated. In decision-making applications, such as autonomous driving or medical diagnosing, the confidence of deep networks plays an important role to bring the trust and reliability to the system. To calibrate the deep networks' confidence, many probabilistic and measure-based approaches are proposed. Temperature Scaling (TS) is a state-of-the-art among
more &raquo; ... asure-based calibration methods which has low time and memory complexity as well as effectiveness. In this paper, we study TS and show it does not work properly when the validation set that TS uses for calibration has small size or contains noisy-labeled samples. TS also cannot calibrate highly accurate networks as well as non-highly accurate ones. Accordingly, we propose Attended Temperature Scaling (ATS) which preserves the advantages of TS while improves calibration in aforementioned challenging situations. We provide theoretical justifications for ATS and assess its effectiveness on wide range of deep models and datasets. We also compare the calibration results of TS and ATS on skin lesion detection application as a practical problem where well-calibrated system can play important role in making a decision.
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1810.11586v3">arXiv:1810.11586v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/dxrllx66orgezifixtoro2l55i">fatcat:dxrllx66orgezifixtoro2l55i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200829142653/https://arxiv.org/pdf/1810.11586v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/d2/be/d2be27ea86b7b4ae227439d0e535f1ce349b246d.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1810.11586v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Unsupervised Temperature Scaling: An Unsupervised Post-Processing Calibration Method of Deep Networks [article]

Azadeh Sadat Mozafari, Hugo Siqueira Gomes, Wilson Leão and Christian Gagné
<span title="2019-06-10">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Correspondence to: Azadeh Sadat Mozafari <azadeh-sadat.mozafari.1@ulaval.ca>. corresponds to the output of a softmax layer, which is typically interpreted as the likeliness (probability) of different class  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1905.00174v3">arXiv:1905.00174v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/dq44sw3yazf7flaxubc434ddte">fatcat:dq44sw3yazf7flaxubc434ddte</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200929193417/https://arxiv.org/pdf/1905.00174v2.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/5d/d0/5dd0e52eeb590ab7ac7e3932136720144a1ba609.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1905.00174v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Biointerface Research in Applied Chemistry: Celebrating the Publication of 600 Papers in 2021, and First Bibliometric Analysis from 2016-2020

<span title="2021-04-17">2021</span> <i title="AMG Transcend Association"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/c2l7bxihmja47gjxjmmghurafa" style="color: black;">Biointerface Research in Applied Chemistry</a> </i> &nbsp;
Mohammed, Aya R. 1 1 Mohareb, Rafat M. 1 1 Mola, Adeleh 1 1 Moradi, Rezvan Seid 1 1 Morata, Antonio 1 1 Moria, Hazim 1 1 Mortazavian, Amir Mohammad 1 1 Mousavi, Sayedali 1 1 Mozafari  ...  Gustavo 1 1 Arteaga-Robalino, Andrea 1 1 Asadzadeh, Sepideh 1 1 Ashyuce, Sevgi 1 1 Asl, Mir Mehdi Chinifroush 1 1 Awadallah, Omima A. 1 1 Ayaad, Dalia M. 1 1 Ayoubi, Mahdi 1 1 Azadeh  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.33263/briac116.1507515140">doi:10.33263/briac116.1507515140</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4vlb7cvs4vbhppzltlealcxugm">fatcat:4vlb7cvs4vbhppzltlealcxugm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210504042753/https://biointerfaceresearch.com/wp-content/uploads/2021/04/20695837116.1507515140.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ae/c9/aec9fe789529e8d47d85ab376eba0fdad7521177.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.33263/briac116.1507515140"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> Publisher / doi.org </button> </a>

Proceedings of "The 16th International and Iranian Congress of Microbiology

Name Sarvenaz, Bigham Soostani, Rahman Patimar, Zarei Darki, Esa Jorjani, Kourosh Sayehmiri, Milad Azami, Zahra Darvishi, Sasan Nikpay, Milad Borji, Pariya Ahmadi Balootaki, Mansour Amin (+746 others)
<span title="">2015</span> <i > Iranian J Publ Health </i> &nbsp; <span class="release-stage">unpublished</span>
TheDETECTION OF SCAF GENE TO IDENTIFYISOLATED STAPHYLOCOCCUS AUREUS Nour Amir Mozafari 1 , Gholam Reza Irajian 1 , *Shiva Mirkalantari 2 1.  ...  Maryam.asadipoor1984@gmail.com : Bordetella pertussis, whole cell, potency, toxicityPERIPLASMIC EXPRESSION AND ONE-STEP PURIFICA- TION OF HEPATITIS B CORE ANTIGEN IN ESCHERICHIA COLI Malihe Sadat  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/3ckr7bt7wre3lhxd7ney6kfmyq">fatcat:3ckr7bt7wre3lhxd7ney6kfmyq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20180427052322/http://ijph.tums.ac.ir/index.php/ijph/article/viewFile/7416/4822" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/21/76/217609a2b0a6c1fdafc83f433a6780b7f4ba8e0d.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a>

Neuronal activity remodels the F-actin based submembrane lattice in dendrites but not axons of hippocampal neurons

Flavie Lavoie-Cardinal, Anthony Bilodeau, Mado Lemieux, Marc-André Gardner, Theresa Wiesner, Gabrielle Laramée, Christian Gagné, Paul De Koninck
<span title="2020-07-20">2020</span> <i title="Springer Science and Business Media LLC"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/tnqhc2x2aneavcd3gx5h7mswhm" style="color: black;">Scientific Reports</a> </i> &nbsp;
Vincent Poiré and Azadeh Sadat Mozafari for preliminary experiments with U-Net implementation and Gabriel Leclerc for assistance with the skeleton analysis.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1038/s41598-020-68180-2">doi:10.1038/s41598-020-68180-2</a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pubmed/32686703">pmid:32686703</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/divhnyj3kffzlpz2chr77nkmsm">fatcat:divhnyj3kffzlpz2chr77nkmsm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201108083305/https://www.nature.com/articles/s41598-020-68180-2.pdf?error=cookies_not_supported&amp;code=34812f7e-cb05-498f-a65a-24ea9668e03f" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/81/90/81905fdbc66afd2f96da9645e2846c4f4e68a492.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1038/s41598-020-68180-2"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> Publisher / doi.org </button> </a>

Local Temperature Scaling for Probability Calibration [article]

Zhipeng Ding, Xu Han, Peirong Liu, Marc Niethammer
<span title="2021-07-27">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Computer methods and programs in biomedicine, 98(3):278-284, 2010. 15 [51] Azadeh Sadat Mozafari, Hugo Siqueira Gomes, Wilson Leão, Steeven Janny, and Christian Gagné.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2008.05105v2">arXiv:2008.05105v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/cqp2fiaqtfdnvcmgdoxguns4qe">fatcat:cqp2fiaqtfdnvcmgdoxguns4qe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210730170033/https://arxiv.org/pdf/2008.05105v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/8a/4a/8a4ade967d9830636cf0160dbde8565a0858edf4.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2008.05105v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>