scholarly journals Structured Output Learning with Conditional Generative Flows

2020 ◽  
Vol 34 (04) ◽  
pp. 5005-5012 ◽  
Author(s):  
You Lu ◽  
Bert Huang
Keyword(s):  
Prediction Models ◽  
State Of The Art ◽  
Prediction Method ◽  
Structured Prediction ◽  
Conditional Likelihood ◽  
Structured Output Learning ◽  
Approximate Likelihood ◽  
Structured Output ◽  
Prediction Problems ◽  
The Relationship

Traditional structured prediction models try to learn the conditional likelihood, i.e., p(y|x), to capture the relationship between the structured output y and the input features x. For many models, computing the likelihood is intractable. These models are therefore hard to train, requiring the use of surrogate objectives or variational inference to approximate likelihood. In this paper, we propose conditional Glow (c-Glow), a conditional generative flow for structured output learning. C-Glow benefits from the ability of flow-based models to compute p(y|x exactly and efficiently. Learning with c-Glow does not require a surrogate objective or performing inference during training. Once trained, we can directly and efficiently generate conditional samples. We develop a sample-based prediction method, which can use this advantage to do efficient and effective inference. In our experiments, we test c-Glow on five different tasks. C-Glow outperforms the state-of-the-art baselines in some tasks and predicts comparable outputs in the other tasks. The results show that c-Glow is versatile and is applicable to many different structured prediction problems.

Parsing Models for Identifying Multiword Expressions

2013 ◽  
Vol 39 (1) ◽  
pp. 195-227 ◽  
Author(s):  
Spence Green ◽  
Marie-Catherine de Marneffe ◽  
Christopher D. Manning
Keyword(s):  
Language Processing ◽  
Prediction Models ◽  
State Of The Art ◽  
Lexical Representation ◽  
Structured Prediction ◽  
Identification Accuracy ◽  
Syntactic Analysis ◽  
Multiword Expressions ◽  
Context Free ◽  
Alternative Theories

Multiword expressions lie at the syntax/semantics interface and have motivated alternative theories of syntax like Construction Grammar. Until now, however, syntactic analysis and multiword expression identification have been modeled separately in natural language processing. We develop two structured prediction models for joint parsing and multiword expression identification. The first is based on context-free grammars and the second uses tree substitution grammars, a formalism that can store larger syntactic fragments. Our experiments show that both models can identify multiword expressions with much higher accuracy than a state-of-the-art system based on word co-occurrence statistics. We experiment with Arabic and French, which both have pervasive multiword expressions. Relative to English, they also have richer morphology, which induces lexical sparsity in finite corpora. To combat this sparsity, we develop a simple factored lexical representation for the context-free parsing model. Morphological analyses are automatically transformed into rich feature tags that are scored jointly with lexical items. This technique, which we call a factored lexicon, improves both standard parsing and multiword expression identification accuracy.

scholarly journals Factors Influencing Creatine Kinase Response in Youth National Team Soccer Players

2021 ◽  
pp. 194173812199938
Author(s):  
Gabor Schuth ◽  
Gyorgy Szigeti ◽  
Gergely Dobreff ◽  
Peter Revisnyei ◽  
Alija Pasic ◽  
...  
Keyword(s):  
Creatine Kinase ◽  
Prediction Models ◽  
Soccer Players ◽  
Linear Regression Models ◽  
Level Of Evidence ◽  
Internal Load ◽  
Youth Soccer ◽  
National Team ◽  
Weekly Training ◽  
The Relationship

Background: Previous studies have examined the relationship between external training load and creatine kinase (CK) response after soccer matches in adults. This study aimed to build training- and match-specific CK prediction models for elite youth national team soccer players. Hypothesis: Training and match load will have different effects on the CK response of elite youth soccer players, and there will be position-specific differences in the most influential external and internal load parameters on the CK response. Study Design: Prospective cohort study. Level of Evidence: Level 4. Methods: Forty-one U16-U17 youth national team soccer players were measured over an 18-month period. Training and match load were monitored with global positioning system devices. Individual CK values were measured from whole blood every morning in training camps. The dataset consisted of 1563 data points. Clustered prediction models were used to examine the relationship between external/internal load and consecutive CK changes. Clusters were built based on the playing position and activity type. The performance of the linear regression models was described by the R2 and the root-mean-square error (RMSE, U/L for CK values). Results: The prediction models fitted similarly during games and training sessions ( R2 = 0.38-0.88 vs 0.6-0.77), but there were large differences based on playing positions. In contrast, the accuracy of the models was better during training sessions (RMSE = 81-135 vs 79-209 U/L). Position-specific differences were also found in the external and internal load parameters, which best explained the CK changes. Conclusion: The relationship between external/internal load parameters and CK changes are position specific and might depend on the type of session (training or match). Morning CK values also contributed to the next day’s CK values. Clinical Relevance: The relationship between position-specific external/internal load and CK changes can be used to individualize postmatch recovery strategies and weekly training periodization with a view to optimize match performance.

scholarly journals Can We Use Grip Strength to Predict Other Types of Hand Exertions? An Example of Manufacturing Industry Workers

2021 ◽  
Vol 18 (3) ◽  
pp. 856
Author(s):  
Victor Ei-Wen Lo ◽  
Yi-Chen Chiu ◽  
Hsin-Hung Tu
Keyword(s):  
Grip Strength ◽  
Manufacturing Industry ◽  
Prediction Models ◽  
Demographic Variable ◽  
Time Saving ◽  
Daily Lives ◽  
Working Environments ◽  
Central Taiwan ◽  
Study Participants ◽  
The Relationship

Background: There are different types of hand motions in people’s daily lives and working environments. However, testing duration increases as the types of hand motions increase to build a normative database. Long testing duration decreases the motivation of study participants. The purpose of this study is to propose models to predict pinch and press strength using grip strength. Methods: One hundred ninety-eight healthy volunteers were recruited from the manufacturing industries in Central Taiwan. The five types of hand motions were grip, lateral pinch, palmar pinch, thumb press, and ball of thumb press. Stepwise multiple linear regression was used to explore the relationship between force type, gender, height, weight, age, and muscle strength. Results: The prediction models developed according to the variable of the strength of the opposite hand are good for explaining variance (76.9–93.1%). Gender is the key demographic variable in the predicting models. Grip strength is not a good predictor of palmar pinch (adjusted-R2: 0.572–0.609), nor of thumb press and ball of thumb (adjusted-R2: 0.279–0.443). Conclusions: We recommend measuring the palmar pinch and ball of thumb strength and using them to predict the other two hand motions for convenience and time saving.

Study on the Prediction Method of Tunnel Surrounding Rock Pressure Based on the Theory of Progressive Destruction

2012 ◽  
Vol 446-449 ◽  
pp. 1432-1436
Author(s):  
Suo Wang
Keyword(s):  
Rock Pressure ◽  
Characteristic Curve ◽  
Prediction Method ◽  
Surrounding Rock ◽  
Progressive Damage ◽  
Tunnel Excavation ◽  
Practical Engineering ◽  
Surrounding Rock Pressure ◽  
The Relationship ◽  
Rock Parameters

In order to predict tunnel surrounding rock pressure, this paper puts forward a series of dynamic numerical simulative model on the tunnel excavation. According to the change of rock damage in the construction program, it adjusts dynamically the mechanical material parameters of surrounding rock. So the model achieves the purpose which is controlling and simulating the process of tunnel progressive damage. In accordance with the numerical simulative results, it analyzes the relationship between the rock parameters with the plastic strain, radial displacement. Then this paper proposes a prediction method of tunnel surrounding rock pressure based on the theory of the progressive damage and method of characteristic curve. Finally, it compares the pressure on the numerical simulative models with on the site date, and it proves that the prediction method has practical engineering value.

The Impact of Stationarity, Regularity, and Context on the Predictability of Individual Human Mobility

2021 ◽  
Vol 7 (4) ◽  
pp. 1-24
Author(s):  
Douglas Do Couto Teixeira ◽  
Aline Carneiro Viana ◽  
Jussara M. Almeida ◽  
Mrio S. Alvim
Keyword(s):  
State Of The Art ◽  
Human Mobility ◽  
Contextual Information ◽  
Prediction Method ◽  
Mobility Prediction ◽  
Mobility Patterns ◽  
Distinct Cell ◽  
Inherent Nature ◽  
The One ◽  
The Impact

Predicting mobility-related behavior is an important yet challenging task. On the one hand, factors such as one’s routine or preferences for a few favorite locations may help in predicting their mobility. On the other hand, several contextual factors, such as variations in individual preferences, weather, traffic, or even a person’s social contacts, can affect mobility patterns and make its modeling significantly more challenging. A fundamental approach to study mobility-related behavior is to assess how predictable such behavior is, deriving theoretical limits on the accuracy that a prediction model can achieve given a specific dataset. This approach focuses on the inherent nature and fundamental patterns of human behavior captured in that dataset, filtering out factors that depend on the specificities of the prediction method adopted. However, the current state-of-the-art method to estimate predictability in human mobility suffers from two major limitations: low interpretability and hardness to incorporate external factors that are known to help mobility prediction (i.e., contextual information). In this article, we revisit this state-of-the-art method, aiming at tackling these limitations. Specifically, we conduct a thorough analysis of how this widely used method works by looking into two different metrics that are easier to understand and, at the same time, capture reasonably well the effects of the original technique. We evaluate these metrics in the context of two different mobility prediction tasks, notably, next cell and next distinct cell prediction, which have different degrees of difficulty. Additionally, we propose alternative strategies to incorporate different types of contextual information into the existing technique. Our evaluation of these strategies offer quantitative measures of the impact of adding context to the predictability estimate, revealing the challenges associated with doing so in practical scenarios.

DESIGNING A HOLISTIC ADAPTIVE RECOMMENDER SYSTEM (HARS) FOR CUSTOMER RELATIONSHIP DEVELOPMENT: A CONCEPTUAL FRAMEWORK

2021 ◽  
Vol IV (2) ◽  
pp. 84-97
Author(s):  
Alina Popa ◽  
Keyword(s):  
Recommender System ◽  
State Of The Art ◽  
Customer Behavior ◽  
Relationship Development ◽  
Customer Relationship ◽  
Limited Data ◽  
Buying Behavior ◽  
Integrative View ◽  
Client Relationship ◽  
The Relationship

With the recent COVID-19 pandemic, the world we knew changed significantly. The buying behavior shifted as well and is reflected by a growing transition to online interaction, higher media consumption and massive turn to online shopping. Companies that aim to remain top of mind to customers should ensure that their way of interacting with user is both relevant and highly adaptive. Companies should invest in state-of-the-art technologies that help manage and optimize the relationship with the client based on both online and offline data. One of the most popular applications that companies use to develop the client relationship is a Recommender System. The vast majority of traditional recommender systems consider the recommendation as a static procedure and focus either on a specific type of recommendation or on some limited data. In this paper, it is proposed a novel Reinforcement Learning-based recommender system that has an integrative view over data and recommendation landscape, as well as it is highly adaptive to changes in customer behavior, the Holistic Adaptive Recommender System (HARS). From system design to detailed activities, it was attempted to present a comprehensive way of designing and developing a HARS system for an e-commerce company use-case as well as giving a suite of metrics that could be used for its evaluation.

scholarly journals Measurement and Prediction of Short-Range Path Loss between 27 and 40 GHz in University Campus Scenarios

2020 ◽  
Author(s):  
Glaucio Ramos ◽  
Carlos Vargas ◽  
Luiz Mello ◽  
Paulo Pereira ◽  
Sandro Gonçalves ◽  
...  
Keyword(s):  
Short Range ◽  
Prediction Models ◽  
Path Loss ◽  
Prediction Method ◽  
University Campus ◽  
Millimetre Wave ◽  
Empirical Prediction ◽  
Loss Prediction ◽  
Path Loss Prediction ◽  
Path Loss Measurements

Abstract In this paper, we present the results of short-range path loss measurements in the microwave and millimetre wave bands, at frequencies between 27 and 40 GHz, obtained in a campaign inside a university campus in Rio de Janeiro, Brazil. Existing empirical path loss prediction models, including the alpha-beta-gamma (ABG) model and the close-in free space reference distance with frequency dependent path loss exponent (CIF) model are tested against the measured data, and an improved prediction method that includes the path loss dependence on the height di erence between transmitter and receiver is proposed. A fuzzy technique is also applied to predict the path loss and the results are compared with those obtained with the empirical prediction models.

scholarly journals MeSHHeading2vec: A new method for representing MeSH headings as feature vectors based on graph embedding algorithm

2019 ◽  
Author(s):  
Zhen-Hao Guo ◽  
Zhu-Hong You ◽  
Hai-Cheng Yi ◽  
Kai Zheng ◽  
Yan-Bin Wang
Keyword(s):  
Computational Models ◽  
Prediction Models ◽  
Graph Embedding ◽  
Computational Prediction ◽  
Additional Information ◽  
Great Hope ◽  
Node Classification ◽  
Relationship Network ◽  
The Relationship ◽  
Mesh Tree

AbstractMotivationEffectively representing the MeSH headings (terms) such as disease and drug as discriminative vectors could greatly improve the performance of downstream computational prediction models. However, these terms are often abstract and difficult to quantify.ResultsIn this paper, we converted the MeSH tree structure into a relationship network and applied several graph embedding algorithms on it to represent these terms. Specifically, the relationship network consisting of nodes (MeSH headings) and edges (relationships) which can be constructed by the rule of tree num. Then, five graph embedding algorithms including DeepWalk (DW), LINE, SDNE, LAP and HOPE were implemented on the relationship network to represent MeSH headings as vectors. In order to evaluate the performance of the proposed method, we carried out the node classification and relationship prediction tasks. The experimental results show that the MeSH headings characterized by graph embedding algorithms can not only be treated as an independent carrier for representation, but also can be utilized as additional information to enhance the distinguishable ability of vectors. Thus, it can act as input and continue to play a significant role in any disease-, drug-, microbe- and etc.-related computational models. Besides, our method holds great hope to inspire relevant researchers to study the representation of terms in this network [email protected]

scholarly journals A Small Sample Prediction Method for Engineering p-S-N Curve

2018 ◽  
Vol 36 (5) ◽  
pp. 831-838 ◽  
Author(s):  
Xiaoran Liu ◽  
Qin Sun ◽  
Ke Liang
Keyword(s):  
Fatigue Life ◽  
Aluminium Alloy ◽  
Stress Level ◽  
Bayesian Method ◽  
Polynomial Chaos ◽  
Least Squares Method ◽  
Prediction Method ◽  
Optimization Method ◽  
Small Sample ◽  
The Relationship

Based on Non-intrusive Polynomial Chaos method, a small sample prediction method for engineering p-S-N curve that has a medium fatigue life is proposed. Parameters in Basquin model are calculated through optimization method based on small sample of observed fatigue life. We establish NIPC polynomials and obtain big sample parameters, obtaining probabilistic properties of parameters with the big sample EDF method. Then the relationship between statistics and stress level are fitted with least squares method. Some new samples are introduced to improve the accuracy of the method. The statistics are updated by Bayesian method. Samples parameters under any stress level are obtained to calculate corresponding fatigue life. Probabilistic properties of fatigue life are predicted, and the p-S-N curve is established. Test observations of aluminium alloy T-2024 are all located in the interval of 95% quantile, showing that the method can effectively predict probabilistic properties of medium fatigue life.

scholarly journals An Efficient Algorithm for Ocean-Front Evolution Trend Recognition

2022 ◽  
Vol 14 (2) ◽  
pp. 259
Author(s):  
Yuting Yang ◽  
Kenneth Kin-Man Lam ◽  
Xin Sun ◽  
Junyu Dong ◽  
Redouane Lguensat
Keyword(s):  
Classification Accuracy ◽  
State Of The Art ◽  
Classification Algorithm ◽  
Final Decision ◽  
Ocean Fronts ◽  
Ocean Front ◽  
Novel Method ◽  
Evolution Laws ◽  
The Relationship ◽  
Evolution Trend

Marine hydrological elements are of vital importance in marine surveys. The evolution of these elements can have a profound effect on the relationship between human activities and marine hydrology. Therefore, the detection and explanation of the evolution laws of marine hydrological elements are urgently needed. In this paper, a novel method, named Evolution Trend Recognition (ETR), is proposed to recognize the trend of ocean fronts, being the most important information in the ocean dynamic process. Therefore, in this paper, we focus on the task of ocean-front trend classification. A novel classification algorithm is first proposed for recognizing the ocean-front trend, in terms of the ocean-front scale and strength. Then, the GoogLeNet Inception network is trained to classify the ocean-front trend, i.e., enhancing or attenuating. The ocean-front trend is classified using the deep neural network, as well as a physics-informed classification algorithm. The two classification results are combined to make the final decision on the trend classification. Furthermore, two novel databases were created for this research, and their generation method is described, to foster research in this direction. These two databases are called the Ocean-Front Tracking Dataset (OFTraD) and the Ocean-Front Trend Dataset (OFTreD). Moreover, experiment results show that our proposed method on OFTreD achieves a higher classification accuracy, which is 97.5%, than state-of-the-art networks. This demonstrates that the proposed ETR algorithm is highly promising for trend classification.

Sign in / Sign up

Export Citation Format

Share Document