13,566 Hits in 8.1 sec

United We Learn Better: Harvesting Learning Improvements From Class Hierarchies Across Tasks [article]

Sindi Shkodrani, Yu Wang, Marco Manfredi, Nóra Baka
2021 arXiv   pre-print
Though ways of best harvesting learning improvements from hierarchies in classification are far from being solved, there is a need to target these problems in other vision tasks such as object detection  ...  In this work we establish a theoretical framework based on probability and set theory for extracting parent predictions and a hierarchical loss that can be used across tasks, showing results across classification  ...  From the results we observe that with leaffocused weighting we harvest improvements over crossentropy on Top-1 accuracy without neglecting the hierarchical metrics, while with hierarchy-focused weighting  ... 
arXiv:2107.13627v1 fatcat:bo7cj54i7fhwbgkkz4vi3vux2i

Building a Better Man Groups for Males [chapter]

2014 Building a Better Man  
Building a Better Man presents a theory and science-based discussion of masculinity in modern America, but it also does much more than that-it interweaves a diverse group of compelling personal stories  ...  Clinicians and group leaders will find that the approach laid out in Building a Better Man leaves clients feeling understood more than judged, which provides a different motivation for change and can set  ...  As we talk about building a better man, we want to learn and repeat the positives from our past, but also grow and learn from the past experiences of others.  ... 
doi:10.4324/9781315886152-12 fatcat:dutfb65skjgg3c4bybcy5crdua

Unlocking Big Data for better health

Steven Munevar
2017 Nature Biotechnology  
We will continue our study of reactions and use this new knowledge to improve chemistry queries across our databases.  ...  Better methods, better science The Goldman group developed a method for inferring cell lineage trees from whole-genome single-cell sequencing data.  ... 
doi:10.1038/nbt.3918 pmid:28700551 fatcat:oqiurm5cgzec7lfenvjvlemxoy

Sustainable, healthy, learning cities and neighbourhoods [chapter]

Sohail Ahmad, University of Glasgow, Gideon Baffoe, Ramjee Bhandari, Graeme Young, Michael Osborne, University of Glasgow, University of Glasgow, University of Glasgow, University of Glasgow
2021 Learning for a Better Future: Perspectives on Higher Education, Cities, Business & Civil Society  
In this scholarly book, the author targets researchers working across various sectors and involved in promoting learning in cities.  ...  The two case studies, which form a substantial part of the chapter, flow from the author's research interest fields before he retired from the  ...  Organisational structures in the future workplace will evolve from the more traditional hierarchies to networks of teams that work across the boundaries of individual organisations and where creative work  ... 
doi:10.4102/aosis.2021.bk214.02 fatcat:aigm5jadaffvjmx5xmgl3ausea

Semantics for interoperability of distributed data and models: Foundations for better-connected information

Ferdinando Villa, Stefano Balbi, Ioannis N. Athanasiadis, Caterina Caracciolo
2017 F1000Research  
semantic specifications that can integrate semantics across diverse domains and disciplines.  ...  We discuss the first outcomes of an investigation in the conceptual and methodological aspects of semantic annotation of data and models, aimed to enable a high standard of interoperability of information  ...  I'm not sure where this requirement is coming from -the user? The task? The worldview?  ... 
doi:10.12688/f1000research.11638.1 fatcat:qtvfe5hao5do3aphmtanlnh5uq

Preparing the ground for better landscape governance: gendered realities in southern Sulawesi

Carol J. Pierce Colfer, Ramadhani Achdiawan, Hasantoha Adnan, Moira Moeliono, Agus Mulyana, Elok Mulyoutami, James M. Roshetko, E. Linda Yuliani, Balang, LepMil
2014 Forests, Trees and Livelihoods  
We conclude with some practical and ethnographically informed suggestions for enhancing collaboration with women and men in these (and similar) communities.  ...  In recognition of the importance of effective and equitable governance at the landscape scale in enhancing human and environmental well-being, we use a recently developed framework for assessing men's  ...  Groups formed from these links might be more effective than existing groups organized by the government, which have had notoriously poor success (cf. the dismal Village Cooperative Units, Koperasi Unit  ... 
doi:10.1080/14728028.2014.951002 fatcat:3a7v5yqfnnaq3aqeqq7fec47wq

Making better use of what we have: Strategies to minimize food waste and resource inefficiency in Canada

Rod MacRae, Anne Siu, Marlee Kohn, Moira Matsubuchi-Shaw, Doug McCallum, Tania Hernandez Cervantes, Danielle Perreault
2016 Canadian Food Studies / La Revue canadienne des études sur l alimentation  
We hope to redress this to some degree in this article.</p>  ...  Such interventions are clearly only a piece of a wide ranging set of initiatives to be undertaken by numerous actors – from food chain firms to individual eaters – but our reading is that more attention  ...  We modified existing food waste reduction hierarchies to both extend our understanding of food systems and create better linkages across the food chain as it relates to waste generation.  ... 
doi:10.15353/cfs-rcea.v3i2.143 fatcat:pzvzoa57xrbmddnzwlx26fvo3u

Community structure models are improved by exploiting taxonomic rank with predictive clustering trees

Jurica Levatić, Dragi Kocev, Marko Debeljak, Sašo Džeroski
2015 Ecological Modelling  
The results of the experimental evaluation reveal that by using the taxonomic rank and the multi-species aspect of the data, we are able to learn better community structure models. (J.  ...  More specifically, we use predictive clustering trees (a generalized form of decision trees) to build models for three practically relevant datasets from the task of community structure modelling: microarthopod  ...  We would also like to acknowledge the support of the European Commission through the project MAESTRA -Learning from Massive, Incompletely annotated, and Structured Data (grant number ICT-2013-612944).  ... 
doi:10.1016/j.ecolmodel.2014.10.023 fatcat:hxmxvjtpk5cpbdllrkax522iqu

Automatic Classification Using DDC on the Swedish Union Catalogue

Koraljka Golub, Johan Hagelbäck, Anders Ardö
2018 International Conference on Theory and Practice of Digital Libraries  
two machine learning algorithms for Swedish catalogue records from the Swedish union catalogue (LIBRIS).  ...  With more and more digital collections of various information resources becoming available, also increasing is the challenge of assigning subject index terms and classes from quality knowledge organization  ...  We are very grateful to Rebecca Green and Sandi Jones for all their advice on how to best process and use the electronic DDC files.  ... 
dblp:conf/ercimdl/GolubHA18 fatcat:t5q7d6ti2bhbbf4a5azldeok6u

Crop mapping from image time series: deep learning with multi-scale label hierarchies [article]

Mehmet Ozgur Turkoglu, Stefano D'Aronco, Gregor Perich, Frank Liebisch, Constantin Streit, Konrad Schindler, Jan Dirk Wegner
2021 arXiv   pre-print
We develop a crop classification method that exploits this expert knowledge and significantly improves the mapping of rare crop types.  ...  This end-to-end trainable, hierarchical network architecture allows the model to learn joint feature representations of rare classes (e.g., apples, pears) at a coarser level (e.g., orchard), thereby boosting  ...  Acknowledgments We thank the Swiss Federal Office for Agriculture (FOAG) for partially funding this Research project through the DeepField Project.  ... 
arXiv:2102.08820v2 fatcat:fl6gxbqqo5b5dgdw3x53jzql4q

Zygarde: Time-Sensitive On-Device Deep Intelligence on Intermittently-Powered Systems [article]

Bashima Islam, Yubo Luo, Shahriar Nirjon
2019 arXiv   pre-print
We develop an imprecise computing-based scheduling algorithm that improves the schedulability of deep learning tasks on intermittently powered systems.  ...  We further propose a semi-supervised machine learning model that exploits the deep features and contributes in determining the imprecise partition of a task.  ...  In Figure 10 , we consider the first two tasks (τ 1 and τ 2 ) from Figure 9(a-b) . We consider a periodic power source with a period of 8 time units.  ... 
arXiv:1905.03854v1 fatcat:rebcm3zjgngvrii6lhne3btvpe

Constrained Semi-supervised Learning in the Presence of Unanticipated Classes

Bhavana Bharat Dalvi
2016 SIGIR Forum  
In this section, we present their hierarchical variants that use a class hierarchy in the learning process to improve over their flat counterparts.  ...  Can a semi-supervised learning method leverage these class constraints to learn better classifiers?  ...  For this task we used the data extracted from the ClueWeb09 corpus [23] by the developers of the NELL KB [91] .  ... 
doi:10.1145/2888422.2888447 fatcat:nqtcg5n5brbvphvh4c3clyrcei

Crop mapping from image time series: Deep learning with multi-scale label hierarchies

Mehmet Ozgur Turkoglu, Stefano D'Aronco, Gregor Perich, Frank Liebisch, Constantin Streit, Konrad Schindler, Jan Dirk Wegner
2021 Remote Sensing of Environment  
We develop a crop classification method that exploits this expert knowledge and significantly improves the mapping of rare crop types.  ...  ZueriCrop covers an area of 50 km × 48 km in the Swiss cantons of Zurich and Thurgau with a total of 116 ′ 000 individual fields spanning 48 crop classes, and 28,000 (multi-temporal) image patches from  ...  Acknowledgments We thank the Swiss Federal Office for Agriculture (FOAG) for partially funding this Research project through the DeepField Project. Appendix A.  ... 
doi:10.1016/j.rse.2021.112603 fatcat:zd7vazk7sbahjabhxxqnpulde4

Multi-Moments in Time: Learning and Interpreting Models for Multi-Action Video Understanding [article]

Mathew Monfort, Bowen Pan, Kandan Ramakrishnan, Alex Andonian, Barry A McNamara, Alex Lascelles, Quanfu Fan, Dan Gutfreund, Rogerio Feris, Aude Oliva
2021 arXiv   pre-print
Here, we present baseline results for multi-action recognition using loss functions adapted for long tail multi-label learning, provide improved methods for visualizing and interpreting models trained  ...  Towards this goal, we present the Multi-Moments in Time dataset (M-MiT) which includes over two million action labels for over one million three second videos.  ...  We can see from the results that pretraining on M-MiT consistently results in better performance on the multi-label datasets.  ... 
arXiv:1911.00232v4 fatcat:u245hymiwjbd7mymr27ukmt4ty

A Hybrid Deep Clustering Approach for Robust Cell Type Profiling Using Single-cell RNA-seq Data: Supplementary Figures and Tables [article]

Suhas Srinivasan, Nathan T Johnson, Dmitry Korkin
2019 bioRxiv   pre-print
It routinely uses machine learning methods, such as feature learning, clustering, and classification, to assist in uncovering novel information from scRNA-seq data.  ...  We also include a technique to estimate an efficient number of latent features in the deep learning model.  ...  Acknowledgements We would like to thank Prof. Eibe Frank for helpful information regarding the functionality of the Weka data mining tool.  ... 
doi:10.1101/511626 fatcat:fp2vlwr63jdsdfv3jdfjqitequ
« Previous Showing results 1 — 15 out of 13,566 results