Filters








58,415 Hits in 3.8 sec

Towards Non-Parametric Learning to Rank [article]

Ao Liu, Qiong Wu, Zhenming Liu, Lirong Xia
2018 arXiv   pre-print
This paper studies a stylized, yet natural, learning-to-rank problem and points out the critical incorrectness of a widely used nearest neighbor algorithm.  ...  Agents rank items nondeterministically according to the Plackett-Luce model, where the higher the utility of an item to the agent, the more likely this item will be ranked high by the agent.  ...  Parametric preference learning has been extensively studied in machine learning, especially learning to rank [11, 31, 32, 4, 7, 6, 5, 29, 20, 17, 39] .  ... 
arXiv:1807.03395v1 fatcat:2povn65u5zbizevfizqlly6uya

Parametric Flatten-T Swish: An Adaptive Non-linear Activation Function For Deep Learning [article]

Hock Hung Chieng, Noorhaniza Wahid, Pauline Ong
2020 arXiv   pre-print
To tackle these shortcomings, this paper introduced Parametric Flatten-T Swish (PFTS) as an alternative to ReLU.  ...  Activation function is a key component in deep learning that performs non-linear mappings between the inputs and outputs.  ...  ELU uses exponential property at the negative region to create a smooth curvature to push the mean activation toward zero.  ... 
arXiv:2011.03155v1 fatcat:neous6r7tnby3fmbjqbncirn7q

PARAMETRIC FLATTEN-T SWISH: AN ADAPTIVE NONLINEAR ACTIVATION FUNCTION FOR DEEP LEARNING

Hock Hung Chieng, Noorhaniza Wahid, Pauline Ong
2020 Journal of Information and Communication Technology  
To tackle these shortcomings, this paper introduced Parametric Flatten-T Swish (PFTS) as an alternative to ReLU.  ...  QActivation function is a key component in deep learning that performs non-linear mappings between the inputs and outputs.  ...  One of the important elements that contribute towards its learning power lies in the non-linear activation function (Ciuparu et al., 2019) .  ... 
doi:10.32890/jict.20.1.2021.9267 fatcat:4tldzw3y3bhobgze7xzrafacze

Non-parametric kernel ranking approach for social image retrieval

Jinfeng Zhuang, Steven C. H. Hoi
2010 Proceedings of the ACM International Conference on Image and Video Retrieval - CIVR '10  
Although the proposed learning scheme is transductive, we suggest some solution to handle unseen data by warping the non-parametric kernel space to some input kernel function.  ...  Unlike existing methods that often adopt some fixed parametric kernel function, our framework learns a non-parametric kernel matrix that can effectively encode the information from both visual and textual  ...  aims to learn an optimized kernel from both textual tags and visual contents of social images; • We present a fast algorithm for non-parametric kernel ranking, which can efficiently learn non-parametric  ... 
doi:10.1145/1816041.1816047 dblp:conf/civr/ZhuangH10 fatcat:nvgulwh54nf6va2a2zcn7v4hdm

The Omani Technical Students' Knowledge of English for Specific Purposes (ESP) Lexis and Their Attitudes Towards Learning ESP

Issa Al Hinai
2018 Sino-US English Teaching  
Nevertheless, little attention has been paid to discover the relationship between the technical students' knowledge of the ESP lexis and their attitudes towards learning ESP in the Omani technical context  ...  Therefore, this research study was conducted to do so and to explore whether the Omani technical students' knowledge of ESP lexis is predictable and is affected by a combination of their attitudes to learning  ...  (two-tailed) 0.000 Table 9 9 Non-Parametric Tests, Mann-Whitney U test: Mean Rank Ranks Gender N Mean rank Sum of ranks Total knowledge score Male 24 13.40 321.50 Female 24 35.60 854.50  ... 
doi:10.17265/1539-8072/2018.10.004 fatcat:zpcv7egcovdcrfuttjoibaip54

Towards Precise Intra-camera Supervised Person Re-identification [article]

Menglin Wang, Baisheng Lai, Haokun Chen, Jianqiang Huang, Xiaojin Gong, Xian-Sheng Hua
2020 arXiv   pre-print
By investigating the characteristics of ICS, this paper proposes camera-specific non-parametric classifiers, together with a hybrid mining quintuplet loss, to perform intra-camera learning.  ...  It is a new setting proposed recently to reduce the burden of annotation while expect to maintain desirable Re-ID performance.  ...  non- parametric classifiers.  ... 
arXiv:2002.04932v2 fatcat:nho3olugd5ez5j5sivzkrrepra

Towards Precise Predictive Modelling of Coronary Artery Disease Elaborating on Omics Data

Georga Elena, Nikolaos Tachos, Antonis Sakellarios, Themis Exarchos, Silvia Rocchiccioli, Gualtiero Pelosi, Operdan Parodi, Lampros Michalis, Dimitrios Fotiadis
2019 Zenodo  
This study aims at developing a patient-specific model for coronary artery disease (CAD) risk stratification based on machine learning modelling of molecular, cellular, inflammatory and omics data.  ...  Three machine learning algorithms, ranging from parametric (i.e. feed-forward neural network) to non-parametric kernel-based ones (i.e. support vector machine) and ensemble models (i.e. random forest),  ...  Variables), and (ii) feature ranking according to the InfoGain criterion (C4).  ... 
doi:10.5281/zenodo.3260258 fatcat:snntd3n5gfeebozilrjv74qsny

The Human Kernel [article]

Andrew Gordon Wilson, Christoph Dann, Christopher G. Lucas, Eric P. Xing
2015 arXiv   pre-print
We use the learned kernels to gain psychological insights and to extrapolate in human-like ways that go beyond traditional stationary and polynomial kernels.  ...  In this paper, we create function extrapolation problems and acquire human responses, and then design a kernel learning framework to reverse engineer the inductive biases of human learners across a set  ...  We also consider empirical non-parametric kernel estimation, since non-parametric kernel estimators can have the flexibility to converge to any positive definite kernel, and thus become appealing when  ... 
arXiv:1510.07389v3 fatcat:gkklb66dl5dztd4fd5mhch4kve

Obtaining Calibrated Probabilities with Personalized Ranking Models [article]

Wonbin Kweon, SeongKu Kang, Hwanjo Yu
2021 arXiv   pre-print
We also design the unbiased empirical risk minimization framework that guides the calibration methods to learning of true preference probability from the biased user-item interaction dataset.  ...  We investigate various parametric distributions and propose two parametric calibration methods, namely Gaussian calibration and Gamma calibration.  ...  Un- biased learning-to-rank with biased feedback.  ... 
arXiv:2112.07428v1 fatcat:pxcngpra3rbc7cq5rbexuf4kmm

Perceptions of teachers, principals and school supervisors on students' skills using ICT in learning in the Senior High School

Desi Rahmatina, Norasykin Mohd Zaid
2019 Malikussaleh Journal of Mathematics Learning (MJML)  
learning, the second to inferential statistical analysis of non parametric statistics Mann Whitney to test alternative hypothesis that there are significant difference skill's students among public school  ...  SPSS software was used to analyze data.The results of the study shown that no different significantly on the students skill on use of computer in learning between public and private school (sig = 0.057  ...  learning, the second to inferential statistical analysis of non parametric statistics Mann Whitney to test alternative hypothesis that there are significant difference skill's students among public school  ... 
doi:10.29103/mjml.v2i1.610 fatcat:gfabqufcpbarjdhnlrwrgzqzfy

Conditioned social dominance threat: observation of others' social dominance biases threat learning

Jan Haaker, Tanaz Molapour, Andreas Olsson
2016 Social Cognitive and Affective Neuroscience  
Participants first learned about the dominance rank of others by observing their dyadic confrontations.  ...  During subsequent fear learning, the dominant and subordinate others were equally predictive of an aversive consequence (mild electric shock) to the participant.  ...  threat learning towards social stimuli.  ... 
doi:10.1093/scan/nsw074 pmid:27217107 pmcid:PMC5040915 fatcat:5zasp4nrkzazpkun75f26cshsi

Students approach towards problem based learning over traditional learning method

Ayesha Juhi
2019 Journal of Medical Science And clinical Research  
The results obtained were compared using non parametric wilcoxon sign rank test.  ...  The approach towards learning that students adopt appears to be an important factor in determining both the quantity and quality of learning.  ...  Non parametric test wilcoxon ranked test for two dependent sample was used to analyse and compare the PBL and LBL score.  ... 
doi:10.18535/jmscr/v7i3.125 fatcat:ehmu5elfxvg2tpv3puvlaivn4y

Teacher's Attitudes towards Educational Technology in English Language Institutes

Narjes Ebrahimi Seraji, Roya Sediq Ziabari, Seyed Jalal Abdolmanafi Rokni
2017 International Journal of English Linguistics  
The non-parametric Spearman Rank-Order Correlation was used to find the relationship between the variables.  ...  There appears to be a positive attitude towards technology, so researchers aimed to seek out new information in an effort to find the relationship among teachers' tenure, age, educational level, experience  ...  Therefore, the non-parametric Spearman Rank-Order Correlation was used to find the relationship. Table 4 below shows the descriptive statistics for the two variables.  ... 
doi:10.5539/ijel.v7n2p176 fatcat:fj3x5fvgtzg63msihfnxzov6em

A STUDY TO EVALUATE USERS' SATISFACTION OF BLACKBOARD LEARN

Md Mokter Hossain, Shakil Akhtar, Muhammad Asadur Rahman
2017 PEOPLE International Journal of Social Sciences  
: http://grdspublishing.org/ significant difference among male vs. female; and engineering vs. non-engineering major users.  ...  This study summarizes and reports first-time users' satisfaction of a Blackboard Learn user interface that had been adopted as an online/blended teaching-learning management tool.  ...  Thus, the non-parametric Wilcoxon-Mann-Whitney U test, which is considered less powerful than the corresponding parametric t-test, was needed to use to analyze the first two research questions (Gay, 2003  ... 
doi:10.20319/pijss.2017.s31.489506 fatcat:trwi4kkm6vgk5d54bfymmta5i4

APRIL: Active Preference Learning-Based Reinforcement Learning [chapter]

Riad Akrour, Marc Schoenauer, Michèle Sebag
2012 Lecture Notes in Computer Science  
In this paper, preference-based reinforcement learning is combined with active ranking in order to decrease the number of ranking queries to the expert needed to yield a satisfactory policy.  ...  Experiments on the mountain car and the cancer treatment testbeds witness that a couple of dozen rankings enable to learn a competent policy.  ...  Along this line, a multipleinstance ranking setting [3] could be used to learn preferences at the fragment (sub-behavior) level, thus making steps toward the definition of sub-behaviors and modular RL  ... 
doi:10.1007/978-3-642-33486-3_8 fatcat:z4lrn4whyzdyxcg6zgies4it7a
« Previous Showing results 1 — 15 out of 58,415 results