11,771 Hits in 6.0 sec

Evaluating Information Retrieval Metrics Based on Bootstrap Hypothesis Tests

Tetsuya Sakai
2007 IPSJ Digital Courier  
We demonstrate the usefulness of our methods using four different data sets (i.e., test collections and submitted runs) from the NTCIR CLIR track series for comparing seven IR metrics, including those  ...  that can handle graded relevance and those based on the Geometric Mean.  ...  Figures 10, 11, 12 and 13 plot, for each IR metric, the Paired/Unpaired Bootstrap ASLs of all system pairs for the NTCIR-3 Chinese/Japanese data.  ... 
doi:10.2197/ipsjdc.3.625 fatcat:l6i64r3dyrcopl2czymxkiusm4

Trial Design and Analysis with Incomplete Paired Data

Mark Chang, Jing Wang
2015 Current Research in Biostatistics  
For a clinical trial design with paired data, it often involves missing observations. In such a case, the data from the trial become a mixture of paired and unpaired data.  ...  We will show how to design classical and adaptive trials with the proposed method. The proposed method can also be used for meta-analysis, in which, some trials with paired data and some are not.  ...  Their method is based on the paired comparison set of responses augmented by a set of missing data indicators for each comparison.  ... 
doi:10.3844/amjbsp.2015.61.68 fatcat:kg4zqnxt2fgzxo6j63vbucryty

Video-based tools to enhance nurses' geriatric knowledge: A development and pilot study

V. Habes, P. Jepma, J.L. Parlevliet, A. Bakker, B.M. Buurman
2020 Nurse Education Today  
The qualitative data showed that contributed to the reflective learning-style and enhanced meaningful learning.  ...  Educational tools that fit the specific learning styles of nurses and nursing students might be useful for this.  ...  Acknowledgements The authors thank Johanneke Helder, Linda van der Voort and Marije Moekotte for performing the interviews and analyzing part of the qualitative data.  ... 
doi:10.1016/j.nedt.2020.104425 pmid:32311666 fatcat:rskmpez4sjazbj7jairwukwuzy

Diversity in Reproductive Health and Human Sexuality: Assessing Attitudes Comfort and Knowledge in Learners Before and After Pilot Curriculum

Rachel Friedlander, Sophie Mou, Lee Shearer
2021 Family Medicine  
We performed unpaired 1-tailed t tests and χ2 tests to compare the scores on the pre- and postcourse surveys. Sample size was 12 students for the first cohort and 23 students for the second cohort.  ...  A disparity between the classroom and virtual setting suggests limitations of online learning for these topics.  ...  Data were pooled rather than paired, as this was a pilot series intended to assess efficacy of the course, and there was concern for high drop-out rates.  ... 
doi:10.22454/fammed.2021.456130 pmid:34019683 fatcat:clwfks62w5h5xdlqhgp2xjhlle

Common Pitfalls in Analysis of Tissue Scores

David K. Meyerholz, Nathan L. Tintle, Amanda P. Beck
2018 Veterinary pathology  
The choice of appropriate statistical test is influenced by the study's experimental design and resultant data (eg, paired vs unpaired, normality, number of groups, etc).  ...  The choice of appropriate statistical test is influenced by the study's experimental design and resultant data (eg, paired vs unpaired, normality, number of groups, etc).  ...  Funding The authors received no financial support for the research, authorship, and/or publication of this article. ORCID iD David K. Meyerholz  ... 
doi:10.1177/0300985818794250 pmid:30131009 fatcat:snzs2con7ng6repqvve42hk7gu

Statistics: a brief overview

Ryan Winters, Andrew Winters, Ronald G Amedee
2010 Ochsner Journal  
The Accreditation Council for Graduate Medical Education sets forth a number of required educational topics that must be addressed in residency and fellowship programs.  ...  , correlation, and numerical versus categorical data.  ...  The test compares the means of 2 data sets to determine if they are equal; if they are, then no difference exists between the sets. It exists as both a paired and unpaired test.  ... 
pmid:21603381 pmcid:PMC3096219 fatcat:bdhif6dlubcndezkpri7buhsiq

Statistical Significance, Power, and Sample Sizes

Tetsuya Sakai
2016 Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval - SIGIR '16  
From those papers that reported enough information for us to conduct power analysis, we identify extremely overpowered and underpowered experiments, as well as appropriate sample sizes for future experiments  ...  Our hope is that this study will help improve the reporting practices and experimental designs of future IR effectiveness studies.  ...  Acknowledgement We thank Professor Hideki Toyoda (Waseda University) for letting us modify his R code and distribute it.  ... 
doi:10.1145/2911451.2911492 dblp:conf/sigir/Sakai16 fatcat:z5dbyh3ufje5vorqi2hjxmq3ly

Designing Test Collections for Comparing Many Systems

Tetsuya Sakai
2014 Proceedings of the 23rd ACM International Conference on Conference on Information and Knowledge Management - CIKM '14  
We demonstrate that, as different evaluation measures have different variances across topics, they inevitably require different topic set sizes.  ...  We provide practical solutions to researchers like her using power analysis and sample size design techniques, and demonstrate its usefulness for several IR tasks and evaluation measures.  ...  Acknowledgement This research is a part of Waseda University's project "Taxonomising and Evaluating Web Search Engine User Behaviours," supported by Microsoft Research.  ... 
doi:10.1145/2661829.2661893 dblp:conf/cikm/Sakai14 fatcat:mm5jqfqwpraldhaz7rk23v7yci

Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes

Song Zhang, Jing Cao, Chul Ahn
2016 Statistics in Medicine  
We propose a hybrid estimator to appropriately account for the mixed nature of observed data: paired outcomes from those who contribute complete pairs of observations and unpaired outcomes from those who  ...  We investigate the estimation of intervention effect and sample size determination for experiments where subjects are supposed to contribute paired binary outcomes with some incomplete observations.  ...  Acknowledgement The work was supported in part by NIH grants 1UL1TR001105, AHRQ grant R24HS22418, CPRIT grants RP110562-C1 and RP120670-C1, and NSF grant IIS-1302497-03.  ... 
doi:10.1002/sim.7168 pmid:27862151 pmcid:PMC5217765 fatcat:honwfilc3jcihbzkkosvgsj2pq

Quantifying effects in two-sample environmental experiments using bootstrap confidence intervals

2007 Environmental Modelling & Software  
Monte Carlo results are discussed and recommendations on data sizes are presented.  ...  case of unpaired experiments, and average of differences and median of differences in case of paired experiments.  ...  Yao (Department of Statistics, London School of Economics, UK) for their comments and advices on aspects of plant physiology and statistics.  ... 
doi:10.1016/j.envsoft.2005.12.001 fatcat:ryp5rip32ffyzjlm5dzvipx4cu

Why we need to report more than 'Data were Analyzed by t-tests or ANOVA'

Tracey L Weissgerber, Oscar Garcia-Valencia, Vesna D Garovic, Natasa M Milic, Stacey J Winham
2018 eLife  
This systematic review examines the quality of reporting for two statistical tests, t-tests and ANOVA, for papers published in a selection of physiology journals in June 2017.  ...  Transparent reporting is essential for the critical evaluation of studies. However, the reporting of statistical methods for studies in the biomedical sciences is often limited.  ...  We do not expect paired data to be negatively correlated -if this happens it is important to review the experimental design and data to ensure that everything is correct.  ... 
doi:10.7554/elife.36163 fatcat:q3wvfctervgtljw6chskvcwjwi

Almost Unsupervised Text to Speech and Automatic Speech Recognition [article]

Yi Ren, Xu Tan, Tao Qin, Sheng Zhao, Zhou Zhao, Tie-Yan Liu
2020 arXiv   pre-print
In this paper, by leveraging the dual nature of the two tasks, we propose an almost unsupervised learning method that only leverages few hundreds of paired data and extra unpaired data for TTS and ASR.  ...  Our method achieves 99.84 level intelligible rate and 2.68 MOS for TTS, and 11.7 dataset, by leveraging only 200 paired speech and text data (about 20 minutes audio), together with extra unpaired speech  ...  Acknowledgement We thank Jun-Wei Gan, Yi Zhuang from Microsoft STC Asia for the further explorations on this work.  ... 
arXiv:1905.06791v3 fatcat:em4vtb7fa5bnfnpylbi4cn42ky

Topic set size design

Tetsuya Sakai
2015 Information retrieval (Boston)  
We employ Nagata's three sample size design techniques, which are based on the paired t test, one-way ANOVA, and confidence intervals, respectively.  ...  These topic set size design methods require topic-by-run score matrices from past test collections for the purpose of estimating the within-system population variance for a particular evaluation measure  ...  Acknowledgments I would like to thank Professor Yasushi Nagata of Waseda University for his valuable advice, and to the guest editors and reviewers for their constructive feedback.  ... 
doi:10.1007/s10791-015-9273-z fatcat:io7hhbty7zhrfhtdclsbvotkji

The Design and Implementation of XiaoIce, an Empathetic Social Chatbot

Li Zhou, Jianfeng Gao, Di Li, Heung-Yeung Shum
2020 Computational Linguistics  
XiaoIce is uniquely designed as an AI companion with an emotional connection to satisfy the human need for communication, affection, and social belonging.  ...  We take into account both intelligent quotient (IQ) and emotional quotient (EQ) in system design, cast human-machine social chat as decision-making over Markov Decision Processes (MDPs), and optimize XiaoIce  ...  The authors are also thankful to colleagues at Microsoft AI & Research for valuable discussions and help with some experiments.  ... 
doi:10.1162/coli_a_00368 fatcat:z67xfwbkbjag3ar6ksm4inbxuu

Advantages of distributed and parallel algorithms that leverage Cloud Computing platforms for large-scale genome assembly

Priti Kumari, Raja Mazumder, Vahan Simonyan, Konstantinos Krampis
2015 F1000Research  
Danio rerio The first phase of the analysis involved a subset of the zebrafish data Results: set (2X coverage) and best results were obtained using K-mer size of 65, while it was observed that Velvet takes  ...  Furthermore, Hadoop clusters can be rented on-demand from Cloud computing providers, and therefore Contrail can provide a simple and cost effective way for genome assembly of data generated at laboratories  ...  using both paired and un-paired sequence data.  ... 
doi:10.12688/f1000research.6016.1 fatcat:bylmprujzjeqxjpokoabolxwpi
« Previous Showing results 1 — 15 out of 11,771 results