Filters








16,771 Hits in 3.8 sec

Probabilistic Compositional Embeddings for Multimodal Image Retrieval [article]

Andrei Neculai, Yanbei Chen, Zeynep Akata
2022 arXiv   pre-print
To learn an informative embedding that can flexibly encode the semantics of various queries, we propose a novel multimodal probabilistic composer (MPC).  ...  Given an arbitrary number of query images and (or) texts, our goal is to retrieve target images containing the semantic concepts specified in multiple multimodal queries.  ...  Given an arbitrary number of multimodal queries (e.g. images and texts), the model is trained to learn a feature embedding for retrieving images that contain the composite set of semantic concepts specified  ... 
arXiv:2204.05845v1 fatcat:ls2rtlso5rfhloopwxcsh4hvcq

Learning User Embeddings from Temporal Social Media Data: A Survey [article]

Fatema Hasan, Kevin S. Xu, James R. Foulds, Shimei Pan
2021 arXiv   pre-print
The temporal nature of user-generated data on social media has largely been overlooked in much of the existing user embedding literature.  ...  In this paper, we survey representative work on learning a concise latent user representation (a.k.a. user embedding) that can capture the main characteristics of a social media user.  ...  dynamics of the embedding at any arbitrary timestamp.  ... 
arXiv:2105.07996v1 fatcat:6elhasieuzce5ogddsl7uukv64

Mixtures of Deep Neural Experts for Automated Speech Scoring

Sara Papi, Edmondo Trentin, Roberto Gretter, Marco Matassoni, Daniele Falavigna
2020 Interspeech 2020  
models, several word embeddings, and two bag-of-word models.  ...  Different deep neural network architectures (both feed-forward and recurrent) are specialized over diverse representations of the texts in terms of: a reference grammar, the outcome of probabilistic language  ...  of an arbitrary number of feature-specific DNNs.  ... 
doi:10.21437/interspeech.2020-1055 dblp:conf/interspeech/PapiTGMF20 fatcat:roiv5qtj5nbe7ksff2ry6ccmnq

Renal Cancer Cell Classification Using Generative Embeddings and Information Theoretic Kernels [chapter]

Manuele Bicego, Aydın Ulaş, Peter Schüffler, Umberto Castellani, Vittorio Murino, André Martins, Pedro Aguiar, Mario Figueiredo
2011 Lecture Notes in Computer Science  
We conclude that the proposed hybrid approach achieves higher accuracy, revealing itself as a promising approach for this class of problems.  ...  In particular we use probabilistic latent semantic analysis (pLSA) as a generative model to perform generative embedding onto the free energy score space (FESS).  ...  In [15] , these IT kernels were successfully used for text categorization, based on multinomial (bag-of-words type) text representations.  ... 
doi:10.1007/978-3-642-24855-9_7 fatcat:ddpq7tj7k5bwpfctxyhn26axla

CLAREL: Classification via retrieval loss for zero-shot learning [article]

Boris N. Oreshkin and Negar Rostamzadeh and Pedro O. Pinheiro and Christopher Pal
2020 arXiv   pre-print
On top of that, we provide a probabilistic justification for a metric rescaling approach that solves a very common problem in the generalized zero-shot learning setting, i.e., classifying test images from  ...  We address the problem of learning fine-grained cross-modal representations. We propose an instance-based deep metric learning approach in joint visual and textual space.  ...  Each batch consists of randomly sampled instances, i.e. pairs of images and their corresponding texts. Images are embedded via ResNet and texts are embedded via a CNN/LSTM stack.  ... 
arXiv:1906.11892v3 fatcat:6xt4awflxrbhbdyk7gu77irh7a

Probabilistic frequent subtrees for efficient graph classification and retrieval

Pascal Welke, Tamás Horváth, Stefan Wrobel
2017 Machine Learning  
We also present different fast techniques for computing the embedding of unseen graphs into (probabilistic frequent) subtree feature spaces.  ...  We also consider partial embeddings, i.e., when only a part of the feature vector has to be calculated.  ...  Acknowledgements The authors thank the anonymous reviewers for their careful reviews and useful suggestions.  ... 
doi:10.1007/s10994-017-5688-7 fatcat:i6qa3j62cvhk5g5nu3mabqgjse

A probabilistic framework for multi-view feature learning with many-to-many associations via neural networks [article]

Akifumi Okuno, Tetsuya Hada, Hidetoshi Shimodaira
2018 arXiv   pre-print
PMvGE is a probabilistic model for predicting new associations via graph embedding of the nodes of data vectors with links of their associations.  ...  A simple framework Probabilistic Multi-view Graph Embedding (PMvGE) is proposed for multi-view feature learning with many-to-many associations so that it generalizes various existing multi-view methods  ...  Therefore, in this paper, we propose a nonlinear framework for multi-view feature learning with manyto-many associations. We name the framework as Probabilistic Multi-view Graph Embedding (PMvGE).  ... 
arXiv:1802.04630v2 fatcat:fu5vb7etsja37d635eopi6hxai

Spherical Paragraph Model [article]

Ruqing Zhang, Jiafeng Guo, Yanyan Lan, Jun Xu, Xueqi Cheng
2017 arXiv   pre-print
To address these problems, we introduce the Spherical Paragraph Model (SPM), a probabilistic generative model based on BoWE, for text representation.  ...  of texts.  ...  All these vMF-based methods treat the text as a single object (i.e., a normalized feature vector), and successfully integrate a directional measure of similarity into a probabilistic setting for text modeling  ... 
arXiv:1707.05635v1 fatcat:7ngupuuw6bcydibc7om3mjhd4q

Learning Continuous User Representations through Hybrid Filtering with doc2vec [article]

Simon Stiebellehner, Jun Wang, Shuai Yuan
2017 arXiv   pre-print
using doc2vec prove to be highly valuable features in supervised machine learning models for look-alike modeling.  ...  In order to maximize the predictive performance of our look-alike modeling algorithms, we propose two novel hybrid filtering techniques that utilize the recent neural probabilistic language model algorithm  ...  The great popularity of neural text embeddings in NLP and the conceptual similarity of generating embeddings from sequences of words and generating embeddings from arbitrary sequences have motivated researching  ... 
arXiv:1801.00215v1 fatcat:avcszeivnjbjxkem6nfe4ctjci

Spherical Paragraph Model [chapter]

Ruqing Zhang, Jiafeng Guo, Yanyan Lan, Jun Xu, Xueqi Cheng
2018 Lecture Notes in Computer Science  
To address these problems, we introduce the Spherical Paragraph Model (SPM), a probabilistic generative model based on BoWE, for text representation.  ...  of texts.  ...  For example, in the popular TF-IDF scheme [11] , each text is represented by tfidf values of selected feature-words.  ... 
doi:10.1007/978-3-319-76941-7_22 fatcat:d2ablbdvqzbsrdvlyuz6z73lku

CompLex: A New Corpus for Lexical Complexity Prediction from Likert Scale Data [article]

Matthew Shardlow, Michael Cooper, Marcos Zampieri
2020 arXiv   pre-print
With a few exceptions, previous studies have approached the task as a binary classification task in which systems predict a complexity value (complex vs. non-complex) for a set of target words in a text  ...  Predicting which words are considered hard to understand for a given target population is a vital step in many NLP applications such as text simplification.  ...  Acknowledgements We would like to thank the anonymous reviewers for their valuable feedback.  ... 
arXiv:2003.07008v3 fatcat:ldjiks6crfcdpou2q3fnbb5h6e

Combining Probabilistic Logic and Deep Learning for Self-Supervised Learning [article]

Hoifung Poon, Hai Wang, Hunter Lang
2021 arXiv   pre-print
These are either added directly (a form of structured self-training) or verified by a human expert (as in feature-based active learning).  ...  We first present deep probabilistic logic(DPL), which offers a unifying framework for task-specific self-supervision by composing probabilistic logic with deep learning.  ...  The latter subsumes feature-based active learning [15] with arbitrary features expressible using probabilistic logic.  ... 
arXiv:2107.12591v1 fatcat:dl2rrtgllvcnhjznpiuerwdynq

Notes on Coalgebras in Stylometry [article]

Joël A. Doat
2021 arXiv   pre-print
In this paper, we discuss how coalgebras are used to formalise the notion of behaviour by embedding syntactic features of a given text into probabilistic transition systems.  ...  By introducing the behavioural distance, we are then able to quantitatively measure differences between points in these systems and thus, comparing features of different texts.  ...  Thus, for every feature, we obtain a behavioural value for this text.  ... 
arXiv:2010.02733v2 fatcat:jxjrcdg3pjet5gmyvx732kdmgy

Using the Naive Bayes as a discriminative classifier [article]

Elie Azeraf, Emmanuel Monfrini, Wojciech Pieczynski
2021 arXiv   pre-print
For classification tasks, probabilistic models can be categorized into two disjoint classes: generative or discriminative.  ...  Moreover, this observation also discusses the notion of Generative-Discriminative pairs, linking, for example, Naive Bayes and Logistic Regression, or HMM and CRF.  ...  In this way, the Naive Bayes can be applied for any classification tasks and consider arbitrary observation features.  ... 
arXiv:2012.13572v3 fatcat:ec22g7qmfnhlzmvjoq666tyi7q

CompLex - A New Corpus for Lexical Complexity Predicition from LikertScale Data

Matthew Shardlow, Marcos Zampieri, Michael Cooper
2020 International Conference on Language Resources and Evaluation  
With a few exceptions, previous studies have approached the task as a binary classification task in which systems predict a complexity value (complex vs. non-complex) for a set of target words in a text  ...  Predicting which words are considered hard to understand for a given target population is a vital step in many NLP applications such as text simplification.  ...  Acknowledgements We would like to thank the anonymous reviewers for their valuable feedback.  ... 
dblp:conf/lrec/ShardlowZC20 fatcat:44bf6arkdvg43kygvhs4ojhzda
« Previous Showing results 1 — 15 out of 16,771 results