Filters








1,378 Hits in 6.5 sec

Hashing as Tie-Aware Learning to Rank [article]

Kun He, Fatih Cakir, Sarah Adel Bargal, Stan Sclaroff
2018 arXiv   pre-print
We first observe that the integer-valued Hamming distance often leads to tied rankings, and propose to use tie-aware versions of AP and NDCG to evaluate hashing for retrieval.  ...  In this paper, we develop learning to rank formulations for hashing, aimed at directly optimizing ranking-based evaluation metrics such as Average Precision (AP) and Normalized Discounted Cumulative Gain  ...  Acknowledgements The authors would like to thank Qinxun Bai, Peter Gacs, and Dora Erdos for helpful discussions.  ... 
arXiv:1705.08562v4 fatcat:uzo3q6h2cranjlmhw7qle7kwlm

Hashing as Tie-Aware Learning to Rank

Kun He, Fatih Cakir, Sarah Adel Bargal, Stan Sclaroff
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
We first observe that the integer-valued Hamming distance often leads to tied rankings, and propose to use tie-aware versions of AP and NDCG to evaluate hashing for retrieval.  ...  In this paper, we develop learning to rank formulations for hashing, aimed at directly optimizing ranking-based evaluation metrics such as Average Precision (AP) and Normalized Discounted Cumulative Gain  ...  Acknowledgements The authors would like to thank Qinxun Bai, Peter Gacs, and Dora Erdos for helpful discussions.  ... 
doi:10.1109/cvpr.2018.00423 dblp:conf/cvpr/0003CBS18 fatcat:x23fwsjwofbg3kt44lxuustjkq

Representation Learning for Efficient and Effective Similarity Search and Recommendation [article]

Casper Hansen
2021 arXiv   pre-print
A common approach is to represent data objects as binary vectors, denoted hash codes, which require little storage and enable efficient similarity search through direct indexing into a hash table or through  ...  State of the art methods use representation learning for generating such hash codes, focusing on neural autoencoder architectures where semantics are encoded into the hash codes by learning to reconstruct  ...  Instead, we compute the tie-aware precision@100 metric [19] corresponding to the average-case retrieval performance.  ... 
arXiv:2109.01815v1 fatcat:tlq2uweeebde5gmi56ubm6rttm

Improving Context Aware Language Models [article]

Aaron Jaech, Mari Ostendorf
2017 arXiv   pre-print
: adaptation of both the hidden and output layers. and a feature hashing bias term to capture context idiosyncrasies.  ...  We show that the most widely-used approach to adaptation (concatenating the context with the word embedding at the input to the recurrent layer) is outperformed by a model that has some low-cost improvements  ...  The setup is similar to our method of using hashing to learn context-dependent biases. However, there are a number of differences.  ... 
arXiv:1704.06380v1 fatcat:xom3fnbitrc5hlhrte7g5l47yu

Discrete Trust-aware Matrix Factorization for Fast Recommendation

Guibing Guo, Enneng Yang, Li Shen, Xiaochun Yang, Xiaodong He
2019 Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence  
Trust-aware recommender systems have received much attention recently for their abilities to capture the influence among connected users.  ...  In this paper we propose a discrete trust-aware matrix factorization (DTMF) model to take dual advantages of both social relations and discrete technique for fast recommendation.  ...  To summarize these before work, learning hash code can be regarded as two independent stages: relaxed learning and binary quantization.  ... 
doi:10.24963/ijcai.2019/191 dblp:conf/ijcai/GuoYSYH19 fatcat:zdqtb7pmszgtxborsgbgpprfom

Building ranked mashups of unstructured sources with uncertain information

Mohamed A. Soliman, Ihab F. Ilyas, Mina Saleeb
2010 Proceedings of the VLDB Endowment  
We present MashRank, a mashup authoring and processing system building on concepts from rank-aware processing, probabilistic databases, and information extraction to enable ranked mashups of (unstructured  ...  MashRank integrates information extraction with query processing by asynchronously pushing extracted data on-the-fly into pipelined rank-aware query plans, and using ranking early-out requirements to limit  ...  ., supervised/unsupervised learning [8, 14]), a limited attention has been given to interleaving extraction with rank-aware query processing, as well as handling extracted results with uncertainty.  ... 
doi:10.14778/1920841.1920947 fatcat:3n2fnluwfjbsxeru2rog2am7dy

Towards Reliable Network Entity Tracking Using Behavioural Bag of Words Representation

Jaroslav Hlavác, Martin Kopp, Michael Polák, Jan Kohout
2021 Conference on Theory and Practice of Information Technologies  
However, normalisation methods traditionally used with BoW in other application domains (e.g. tf-idf, stop words) do not work well with network data as they are not designed to capture behavioral patterns  ...  Our results show that using multiple data sources significantly improves entity tracking, especially when combined with proposed time-aware normalisation.  ...  It brings the possibility to tie the (dis)similarity of two objects with concrete features.  ... 
dblp:conf/itat/HlavacKPK21 fatcat:q4y3t5ccnzavdbt5u5ygtmc57e

A Survey on Deep Hashing Methods [article]

Xiao Luo, Haixin Wang, Daqing Wu, Chong Chen, Minghua Deng, Jianqiang Huang, Xian-Sheng Hua
2022 arXiv   pre-print
Specifically, we categorize deep supervised hashing methods into pairwise methods, ranking-based methods, pointwise methods as well as quantization according to how measuring the similarities of the learned  ...  learning manners.  ...  HashMI [8] - Mutual Information - Drop - TALR [59] - Relaxed AP + NDCG - Tie-Awareness MLRDH [117] - - Multi-linear Reg. Alternation Hash Boosting HCBDH [25] - - Cla.  ... 
arXiv:2003.03369v5 fatcat:m2iu3htilvgztkcazw3cyk6iqe

2020 Index IEEE Transactions on Image Processing Vol. 29

2020 IEEE Transactions on Image Processing  
., +, TIP 2020 72-84 From Rank Estimation to Rank Approximation: Rank Residual Constraint for Image Restoration.  ...  Ko, C., +, TIP 2020 6918-6931 From Rank Estimation to Rank Approximation: Rank Residual Constraint for Image Restoration.  ... 
doi:10.1109/tip.2020.3046056 fatcat:24m6k2elprf2nfmucbjzhvzk3m

Adopting the cybersecurity concepts into curriculum: the potential effects on students' cybersecurity knowledge

Mohammad Azzeh, Ahmad Mousa Altamimi, Mahmood Albashayreh, Mohammad A AL-Oudat
2022 Indonesian Journal of Electrical Engineering and Computer Science  
To this end, a pilot study was first conducted to measure the current level of cybersecurity awareness. The results revealed that students do not have much knowledge of cybersecurity.  ...  examines the effect of adopting cybersecurity concepts on the information and technology (IT) curriculum and determines the potential effect on students' knowledge of cybersecurity practices and level of awareness  ...  ACKNOWLEDGEMENTS The authors are grateful to the Applied Science Private University, Amman-Jordan, for the full financial support granted to cover the publication fee of this research article.  ... 
doi:10.11591/ijeecs.v25.i3.pp1749-1758 fatcat:hhs6zyaabvby3lm74ilcc6rffy

Isometric Graph Neural Networks [article]

Matthew Walker, Bo Yan, Yiou Xiao, Yafei Wang, Ayan Acharya
2020 arXiv   pre-print
In addition to an improvement in AUC-ROC as high as 43% in these experiments, we observe a consistent and substantial improvement as high as 400 measure that directly reflects distance information, demonstrating  ...  To enable this highly desired capability, we propose a technique to learn Isometric Graph Neural Networks (IGNN), which requires changing the input representation space and loss function to enable any  ...  A tie that occurs for a pair in both A and B is not counted in either T or U . We compare the rankings in the ascending order of graph distance and in the descending order of cosine similarity.  ... 
arXiv:2006.09554v1 fatcat:jytiaunwt5aqhaeuj4u5o6y42e

Big Data-Driven Marketing: How Machine Learning Outperforms Marketers' Gut-Feeling [chapter]

Pål Sundsøy, Johannes Bjelland, Asif M. Iqbal, Alex "Sandy" Pentland, Yves-Alexandre de Montjoye
2014 Lecture Notes in Computer Science  
Using historical data, a machine learning prediction model is then trained, validated, and used to select a treatment group.  ...  We present results from a large-scale experiment in a MNO in Asia where we use machine learning to segment customers for text-based marketing.  ...  The conclusions in this document are those of the authors and should not be interpreted as representing the social policies, either expressed or implied, of the sponsors  ... 
doi:10.1007/978-3-319-05579-4_45 fatcat:nb34sqyrkrbonaz7fszs2ev6ri

Fast and Memory-Efficient Neural Code Completion [article]

Alexey Svyatkovskiy, Sebastian Lee, Anna Hadjitofi, Maik Riechert, Juliana Franco, Miltiadis Allamanis
2021 arXiv   pre-print
This allows us to explore the design space and evaluate different techniques.  ...  While deep learning has made significant progress in the statistical prediction of source code, state-of-the-art neural network models consume hundreds of megabytes of memory, bloating the development  ...  As with all hashing methods, this introduces collisions as multiple features may have the same hash, which nevertheless the machine learning models can learn to overcome to some extent.  ... 
arXiv:2004.13651v4 fatcat:ip2kpfsc5fcqdog5uzq7f6ps3u

Cost-aware Feature Selection for IoT Device Classification [article]

Biswadeep Chakraborty, Dinil Mon Divakaran, Ido Nevat, Gareth W. Peters, Mohan Gurusamy
2020 arXiv   pre-print
We define and formulate the problem of cost-aware IoT device classification.  ...  Recent works have explored machine learning techniques for fingerprinting (or classifying) IoT devices, with promising results.  ...  The memory requirement is assigned as high due to the necessity to maintain a hash table.  ... 
arXiv:2009.01368v2 fatcat:focyniynrfchdel4z4dd3ybcmi

A survey of query log privacy-enhancing techniques from a policy perspective

Alissa Cooper
2008 ACM Transactions on the Web  
A user control is defined as a mechanism that allows individual Internet users to choose to have the technique applied to their own query logs.  ...  As popular search engines face the sometimes conflicting interests of protecting privacy while retaining query logs for a variety of uses, numerous technical measures have been suggested to both enhance  ...  ACKNOWLEDGMENTS Many thanks to my colleagues Jim Dempsey, John Morris, Ari Schwartz, David Sohn, Peter Swire, and Paul Otto for their helpful feedback in developing this analysis.  ... 
doi:10.1145/1409220.1409222 fatcat:dhht2jowkra73o2ppgbmxu3jke
« Previous Showing results 1 — 15 out of 1,378 results