scholarly journals Task Embedded Coordinate Update: A Realizable Framework for Multivariate Non-Convex Optimization

2019 ◽  
Vol 33 ◽  
pp. 1650-1657
Author(s):  
Yiyang Wang ◽  
Risheng Liu ◽  
Long Ma ◽  
Xiaoliang Song
Keyword(s):  
Numerical Algorithms ◽  
Coordinate Descent ◽  
Unified Framework ◽  
Convex Problems ◽  
Efficiency And Effectiveness ◽  
Task Oriented ◽  
To Receive ◽  
Coordinate Update ◽  
Embedded Strategies ◽  
Theoretical Analyses

We in this paper propose a realizable framework TECU, which embeds task-specific strategies into update schemes of coordinate descent, for optimizing multivariate non-convex problems with coupled objective functions. On one hand, TECU is capable of improving algorithm efficiencies through embedding productive numerical algorithms, for optimizing univariate sub-problems with nice properties. From the other side, it also augments probabilities to receive desired results, by embedding advanced techniques in optimizations of realistic tasks. Integrating both numerical algorithms and advanced techniques together, TECU is proposed in a unified framework for solving a class of non-convex problems. Although the task embedded strategies bring inaccuracies in sub-problem optimizations, we provide a realizable criterion to control the errors, meanwhile, to ensure robust performances with rigid theoretical analyses. By respectively embedding ADMM and a residual-type CNN in our algorithm framework, the experimental results verify both efficiency and effectiveness of embedding task-oriented strategies in coordinate descent for solving practical problems.

scholarly journals UniMF: A Unified Framework to Incorporate Multimodal Knowledge Bases intoEnd-to-End Task-Oriented Dialogue Systems

2021 ◽  
Author(s):  
Shiquan Yang ◽  
Rui Zhang ◽  
Sarah M. Erfani ◽  
Jey Han Lau
Keyword(s):  
Large Scale ◽  
Knowledge Bases ◽  
Language Models ◽  
Dialogue Systems ◽  
Unified Framework ◽  
Proposed Model ◽  
Multimodal Information ◽  
Task Oriented ◽  
Novel Model ◽  
Single Modality

Knowledge bases (KBs) are usually essential for building practical dialogue systems. Recently we have seen rapidly growing interest in integrating knowledge bases into dialogue systems. However, existing approaches mostly deal with knowledge bases of a single modality, typically textual information. As today's knowledge bases become abundant with multimodal information such as images, audios and videos, the limitation of existing approaches greatly hinders the development of dialogue systems. In this paper, we focus on task-oriented dialogue systems and address this limitation by proposing a novel model that integrates external multimodal KB reasoning with pre-trained language models. We further enhance the model via a novel multi-granularity fusion mechanism to capture multi-grained semantics in the dialogue history. To validate the effectiveness of the proposed model, we collect a new large-scale (14K) dialogue dataset MMDialKB, built upon multimodal KB. Both automatic and human evaluation results on MMDialKB demonstrate the superiority of our proposed framework over strong baselines.

An Approach to Capturing Balance between Retrieval Effectiveness and Efficiency

2011 ◽  
Vol 268-270 ◽  
pp. 1773-1779
Author(s):  
Ke Xu ◽  
Long Fei Li
Keyword(s):  
User Satisfaction ◽  
Learning To Rank ◽  
Loss Functions ◽  
Query Execution ◽  
Ranking Functions ◽  
Production Environment ◽  
Unified Framework ◽  
Retrieval Effectiveness ◽  
Efficiency And Effectiveness ◽  
Effectiveness And Efficiency

It has been shown that learning to rank approaches are capable of learning highly effective ranking functions. However, these approaches have mostly ignored the important issue of efficiency. Given that both efficiency and effectiveness are important for real search engines, models that are optimized for effectiveness may not meet the strict efficiency requirements necessary to deploy in a production environment. In this work, we present a unified framework for jointly optimizing effectiveness and efficiency. We propose new metrics that capture the tradeoff between these two competing forces and devise a strategy for automatically learning models that directly optimize the tradeoff metrics. Experiments indicate that models learned in this way provide a good balance between retrieval effectiveness and efficiency. With specific loss functions, learned models converge to familiar existing ones, which demonstrate the generality of our framework. Finally, we show that our approach naturally leads to a reduction in the variance of query execution times, which is important for query load balancing and user satisfaction.

scholarly journals Combining rTMS and Task-Oriented Training in the Rehabilitation of the Arm after Stroke: A Pilot Randomized Controlled Trial

2013 ◽  
Vol 2013 ◽  
pp. 1-8 ◽  
Author(s):  
Johanne Higgins ◽  
Lisa Koski ◽  
Haiqun Xie
Keyword(s):  
Randomized Controlled Trial ◽  
Cortical Excitability ◽  
Repetitive Transcranial Magnetic Stimulation ◽  
Controlled Trial ◽  
Preliminary Evidence ◽  
Randomized Controlled ◽  
Arm Function ◽  
Task Oriented ◽  
To Receive ◽  
General Community

Introduction. Repetitive transcranial magnetic stimulation (rTMS) is a promising technique for promoting rehabilitation of arm function after stroke. The feasibility and impact of rTMS as an adjunct to traditional task-oriented training to improve arm function have not yet been demonstrated.Objective. Evaluate the feasibility of a randomized controlled trial aimed at determining the efficacy of rTMS as an adjunct to task-oriented therapy in facilitating restoration of arm function after stroke.Methods. Stratified block-randomized controlled trial set in the general community. Eleven stroke persons with mild to severe arm deficits were recruited and randomized to receive 8 sessions ofreal-rTMS orsham-rTMS followed by ninety minutes of arm tasks designed to improve function.Results. Medium to large, statistically significant effect sizes (0.49 to 1.63) were observed in both groups on several measures of arm function at the postintervention evaluation. Three out of four subjects in thereal-TMS condition showed increased levels of corticomotor excitability after the first stimulation session.Conclusions. Preliminary evidence suggests that an rTMS protocol potent enough to induce transient increases in cortical excitability of the lesioned hemisphere is feasible but did not show promising results as an adjunct to task-specific training. This trial is registration with Clinical Trials.govNCT00850408.

scholarly journals The Contribution of the EC Energy Certificate in Improving Sustainability of the Housing Stock

10.1068/b3118 ◽  
2005 ◽  
Vol 32 (1) ◽  
pp. 21-31 ◽  
Author(s):  
Milou Beerepoot ◽  
Minna Sunikka
Keyword(s):  
Energy Savings ◽  
Housing Stock ◽  
Energy Performance ◽  
Urban Housing ◽  
Innovative Products ◽  
Efficiency And Effectiveness ◽  
Gradual Development ◽  
Interesting Approach ◽  
To Receive ◽  
New Buildings

In 2003 the European Commission introduced the EC Directive on the energy performance of buildings in recognition of the importance of energy savings in the urban housing stock. The Directive gives the member states freedom to design the different elements in practice. The energy certificate for existing buildings demanded by the EC Directive can be used as a communicative instrument, or combined with economic or regulatory principles. The authors discuss the anticipated efficiency and effectiveness of different policy approaches in the application of the EC energy certificate for the urban housing stock. They argue that, although energy certificates as a communication instrument for household appliances have appeared to be relatively successful, the different nature of the building sector may mean that their effectiveness here will be rather limited. The combination of energy certificates with tax schemes seems promising, but will have to be coupled with general income taxes or in housing-related taxes in order to prevent regressive social effects. Combination of the energy certificate with subsidies should be limited, because of the ‘free-rider effect’, and subsidies should only cover innovative products at the beginning of their ‘learning curve’. Effective results can probably be expected from the introduction of regulations combined with energy-certificate standards, but this requires a rather drastic approach and needs time to receive sufficient commitment, as has been the case for new buildings where there has been a gradual development of energy regulations over the last thirty years. However, an introduction of energy standards for the existing urban housing stock through the EC energy certificate offers great potential in the realisation of CO2 reductions. The introduction of an energy standard, by means of the energy certificate in combination with progressive taxes or other economic measures to reward better and punish worse energy-performance levels, seems an interesting approach that needs further research.

scholarly journals Analisis Kinerja Manajemen PT Sariguna Primatirta Tbk. Menggunakan Metode Balanced Scorecard Periode 2018-2020

2021 ◽  
Vol 4 (2) ◽  
pp. 37-50
Author(s):  
Mohammad Khoiruzi Afiq ◽  
Helmina Ardyanfitri
Keyword(s):  
Drinking Water ◽  
Performance Appraisal ◽  
Balanced Scorecard ◽  
Quantitative Research ◽  
Safety Management ◽  
Management Performance ◽  
The Balanced Scorecard ◽  
Efficiency And Effectiveness ◽  
The Government ◽  
To Receive

The high demand for quality drinking water in Indonesia, at least Indonesia needs 570.40 million liters every day. PT Sariguna Primatirta Tbk is a drinking water producer with the first producer in Indonesia to receive a food safety management ISO 22000: 2005 certificate and has a Top Brand Index 2021 achievement of 39.7% with the first rank in the Oxygenated Bottled Drinking Water category. Management performance appraisal is important not only for internal, but also for the government, investors and the general public. This assessment can be used as a tool used to assess the level of efficiency and effectiveness of management in implementing strategies and preventing company losses. This study aims to determine the management performance of PT Sariguna Primatirta Tbk for the 2018-2020 period using the Balanced Scorecard method. The method used in this research is descriptive quantitative research. The object of this research is the Financial Statements of PT Sariguna Primatrita Tbk for 2018-2020. The results of this study indicate that the performance of PT Sariguna Primatirta Tbk in 2018 to 2019 is categorized as good by getting a total score of 16. However, in 2019 to 2020 there was a decline in performance due to non-optimal use of all assets to generate company profits but still got a total score of 10.

scholarly journals La Reforma Educativa de 2013 y sus efectos en la precarización del trabajo docente en México

2020 ◽  
Vol 28 ◽  
pp. 140
Author(s):  
Cesar Silva Montes ◽  
Arturo Gutiérrez Lozano
Keyword(s):  
Public Education ◽  
Working Conditions ◽  
Basic Education ◽  
Teaching Profession ◽  
Ciudad Juarez ◽  
Labor Rights ◽  
Efficiency And Effectiveness ◽  
Reforma Educativa ◽  
Stable Employment ◽  
To Receive

This article analyzes the increase in the precariousness of Mexican public education since the approval and implementation of the Educational Reform in 2013. The basis of the study is the analysis of the discourse of public basic education teachers in Ciudad Juarez regarding the loss of earned labor rights. For example, the inability of new teachers to receive the same salaries and benefits as the previous generation of teachers, as well as the increase in responsibilities, activities, and obligations. In addition, teaching flexibilization allows teachers to recognize working conditions as precarious, compared to previous jobs, they conceive it as stable employment. This research shows how managerialism, Toyotism, flexibilization and outsourcing models according to the international neoliberal trend eliminates the union identity of the teaching profession and reconfigures it on the basis of the logic of efficiency and effectiveness through standardized and decontextualized evaluations, as a recent phenomenon in the Mexican teaching profession.

scholarly journals A Coordinate Descent Method for Total Variation Minimization

2017 ◽  
Vol 2017 ◽  
pp. 1-13 ◽  
Author(s):  
Hong Deng ◽  
Dongwei Ren ◽  
Gang Xiao ◽  
David Zhang ◽  
Wangmeng Zuo
Keyword(s):  
Total Variation ◽  
Augmented Lagrangian Method ◽  
Optimal Solution ◽  
Descent Method ◽  
Coordinate Descent ◽  
Total Variation Minimization ◽  
Efficiency And Effectiveness ◽  
Weighted Median ◽  
Small Random Perturbations ◽  
New Perspective

Total variation (TV) is a well-known image model with extensive applications in various images and vision tasks, for example, denoising, deblurring, superresolution, inpainting, and compressed sensing. In this paper, we systematically study the coordinate descent (CoD) method for solving general total variation (TV) minimization problems. Based on multidirectional gradients representation, the proposed CoD method provides a unified solution for both anisotropic and isotropic TV-based denoising (CoDenoise). With sequential sweeping and small random perturbations, CoDenoise is efficient in denoising and empirically converges to optimal solution. Moreover, CoDenoise also delivers new perspective on understanding recursive weighted median filtering. By incorporating with the Augmented Lagrangian Method (ALM), CoD was further extended to TV-based image deblurring (ALMCD). The results on denoising and deblurring validate the efficiency and effectiveness of the CoD-based methods.

scholarly journals AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search

2020 ◽  
Author(s):  
Daoyuan Chen ◽  
Yaliang Li ◽  
Minghui Qiu ◽  
Zhen Wang ◽  
Bofang Li ◽  
...  
Keyword(s):  
Language Processing ◽  
Language Models ◽  
Compression Method ◽  
Neural Architecture ◽  
Efficiency And Effectiveness ◽  
Comparable Performance ◽  
Knowledge Distillation ◽  
Good Trade ◽  
Task Oriented ◽  
Small Models

Large pre-trained language models such as BERT have shown their effectiveness in various natural language processing tasks. However, the huge parameter size makes them difficult to be deployed in real-time applications that require quick inference with limited resources. Existing methods compress BERT into small models while such compression is task-independent, i.e., the same compressed BERT for all different downstream tasks. Motivated by the necessity and benefits of task-oriented BERT compression, we propose a novel compression method, AdaBERT, that leverages differentiable Neural Architecture Search to automatically compress BERT into task-adaptive small models for specific tasks. We incorporate a task-oriented knowledge distillation loss to provide search hints and an efficiency-aware loss as search constraints, which enables a good trade-off between efficiency and effectiveness for task-adaptive BERT compression. We evaluate AdaBERT on several NLP tasks, and the results demonstrate that those task-adaptive compressed models are 12.7x to 29.3x faster than BERT in inference time and 11.5x to 17.0x smaller in terms of parameter size, while comparable performance is maintained.

Accuracy Estimation of the Approximated Methods Used for Assessing Risk of Buildings Damage Under the Influence of Underground Exploitation in the Light of World’s and Polish Experience – Part 2 / Analiza Dokładności Przybliżonych Metod Oceny Zagrożenia Budynków Wpływami Podziemnej Eksploatacji Stosowanych W Świecie I Polskiej Metody Punktowej – Część 2

2013 ◽  
Vol 58 (3) ◽  
pp. 855-865 ◽  
Author(s):  
Agnieszka Malinowska
Keyword(s):  
Hard Coal ◽  
Accuracy Estimation ◽  
Polish Experience ◽  
Efficiency And Effectiveness ◽  
The One ◽  
Theoretical Analyses ◽  
Damage Risk

Abstract This paper is a continuation of theoretical analyses of World’s methods used for assessing damage risk to buildings with continuous strains, which were presented in Part 1. The authors focus only on those methods in which the scale of damage to buildings can be approximated. Selected methods were tested on 100 random objects sited in hard coal excavation-induced areas. The efficiency and effectiveness of those methods was evaluated. The damage risk was also verified with the use of a method currently used in Poland. The efficiency results obtained for World’s methods and the one used in Poland turned out to be comparable. Practical studies were made to evaluate the adaptability of those methods in the underground exploitation-induced conditions in Poland.

Sign in / Sign up

Export Citation Format

Share Document