scholarly journals Understanding VAEs in Fisher-Shannon Plane

Author(s):  
Huangjie Zheng ◽  
Jiangchao Yao ◽  
Ya Zhang ◽  
Ivor W. Tsang ◽  
Jia Wang

In information theory, Fisher information and Shannon information (entropy) are respectively used to quantify the uncertainty associated with the distribution modeling and the uncertainty in specifying the outcome of given variables. These two quantities are complementary and are jointly applied to information behavior analysis in most cases. The uncertainty property in information asserts a fundamental trade-off between Fisher information and Shannon information, which enlightens us the relationship between the encoder and the decoder in variational auto-encoders (VAEs). In this paper, we investigate VAEs in the Fisher-Shannon plane, and demonstrate that the representation learning and the log-likelihood estimation are intrinsically related to these two information quantities. Through extensive qualitative and quantitative experiments, we provide with a better comprehension of VAEs in tasks such as high-resolution reconstruction, and representation learning in the perspective of Fisher information and Shannon information. We further propose a variant of VAEs, termed as Fisher auto-encoder (FAE), for practical needs to balance Fisher information and Shannon information. Our experimental results have demonstrated its promise in improving the reconstruction accuracy and avoiding the noninformative latent code as occurred in previous works.

Author(s):  
Paolo FESTA ◽  
Tommaso CORA ◽  
Lucilla FAZIO

Is it possible to transform stone into a technological and innovative device? The meeting with one of the main stone transformers in Europe produced the intention of a disruptive operation that could affect the strategy of the whole company. A contagious singularity. By intertwining LEAN methodologies and the human-centric approach of design thinking, we mapped the value creation in the company activating a dialogue with the workers and the management, listening to people, asking for ambitions, discovering problems and the potential of production. This qualitative and quantitative analysis conducted with a multidisciplinary approach by designers, architects and marketing strategists allowed us to define a new method. We used it to design a platform that could let all the players express their potential to the maximum. This is how the group's research laboratory was born, with the aim of promoting the relationship between humans and stone through product innovation. With this goal, we coordinated the new team, developing technologies that would allow creating a more direct relationship between man and surface, making the stone reactive. The result was the first responsive kitchen ever.


Psych ◽  
2021 ◽  
Vol 3 (2) ◽  
pp. 197-232
Author(s):  
Yves Rosseel

This paper discusses maximum likelihood estimation for two-level structural equation models when data are missing at random at both levels. Building on existing literature, a computationally efficient expression is derived to evaluate the observed log-likelihood. Unlike previous work, the expression is valid for the special case where the model implied variance–covariance matrix at the between level is singular. Next, the log-likelihood function is translated to R code. A sequence of R scripts is presented, starting from a naive implementation and ending at the final implementation as found in the lavaan package. Along the way, various computational tips and tricks are given.


Stats ◽  
2019 ◽  
Vol 2 (2) ◽  
pp. 247-258 ◽  
Author(s):  
Pedro L. Ramos ◽  
Francisco Louzada

A new one-parameter distribution is proposed in this paper. The new distribution allows for the occurrence of instantaneous failures (inliers) that are natural in many areas. Closed-form expressions are obtained for the moments, mean, variance, a coefficient of variation, skewness, kurtosis, and mean residual life. The relationship between the new distribution with the exponential and Lindley distributions is presented. The new distribution can be viewed as a combination of a reparametrized version of the Zakerzadeh and Dolati distribution with a particular case of the gamma model and the occurrence of zero value. The parameter estimation is discussed under the method of moments and the maximum likelihood estimation. A simulation study is performed to verify the efficiency of both estimation methods by computing the bias, mean squared errors, and coverage probabilities. The superiority of the proposed distribution and some of its concurrent distributions are tested by analyzing four real lifetime datasets.


2019 ◽  
Vol 6 (1) ◽  
pp. 191-197 ◽  
Author(s):  
John B. Rundle ◽  
Alexis Giguere ◽  
Donald L. Turcotte ◽  
James P. Crutchfield ◽  
Andrea Donnellan

Author(s):  
T.C. Macgregor ◽  
J.N Nwaiwu

But knowing the unknown and therefore estimating the relationship between accounting information quality and corporate performance are still a difficult task. The aim of this empirical study is to explore the relationship between the accounting information quality and corporate performance of oil and gas companies in Nigeria. Data on different types of accounting information quality and return on equity were primarily collected from the respondents and analyzed using ordinary least square regression analysis the data with the aid of statistical package for social sciences version 25.0. The empirical result indicates that accounting information quality significantly relate to return n equity; explaining about 85.1% of the total variation in return and equity. Relevance, faithful representation was each found to significantly relate to return on equity. The study empirically conclude that accounting information quality has the potency to make significant contribution to quoted financial performance of oil and gas companies and recommends that having investigated theoretical and empirical issues, also considering the findings and conclusion, the following recommendations were made: There should be need for preparers of accounting information to improve on the accounting information quality devoid of window dressing and creative accounting, regular disclosure, transparency and accountability of such accounting information is required since investors are sensitive to qualitative and quantitative accounting information in assessing the performance of quoted oil companies in and outside Nigeria. Also in line with qualitative and quantitative of accounting information quality, financial statements of quoted oil companies in Nigeria should be prepared and presented according to laid down regulations and ethical standards duly observed to ensure accounting information presented for among users, most and public consumption do represent the oil companies’ economic reality during reported periods.


2021 ◽  
Vol 6 (16) ◽  
pp. 181-188
Author(s):  
Helmi Hamzah ◽  
Noriah Othman ◽  
Norainiratna Badrulhisham ◽  
Lina Karlinasari

Urban trees are exposed to "unintentional vandalism" during poorly skilled pruning practices that can lead to tree structure damage. This causes harmful consequences that affect tree performance in terms of the ecosystem services they contribute. This study aims to explore the relationship between unintentional tree vandalism and poorly skilled pruning practices in tree maintenance by analysing qualitative and quantitative data from tree workers contracted by selected Malaysian local authorities and tree care experts. The results showed that insufficient tree pruning knowledge leads to unintentional vandalism in tree pruning practices. Keywords: Tree management; unskilled tree pruning; unintentional tree vandalism; urban tree eISSN: 2398-4287© 2021. The Authors. Published for AMER ABRA cE-Bs by e-International Publishing House, Ltd., UK. This is an open access article under the CC BYNC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer–review under responsibility of AMER (Association of Malaysian Environment-Behaviour Researchers), ABRA (Association of Behavioural Researchers on Asians/Africans/Arabians) and cE-Bs (Centre for Environment-Behaviour Studies), Faculty of Architecture, Planning & Surveying, Universiti Teknologi MARA, Malaysia. DOI: https://doi.org/10.21834/ebpj.v6i16.2633


2020 ◽  
Vol 6 (2) ◽  
pp. 72-84
Author(s):  
Endro Tri Susdarwono

The purpose of this study discusses how mastery of 4 (four) basic prerequisites of arithmetic which includes the ability to count, make Arithmetic / multiples, complement especially Nines and ten, and the concept of place values in numbers affect students' ability to solve math problems consisting of addition, subtraction, multiplication, and multiplication. The approach in this study uses a descriptive approach, the method used is a combination of qualitative and quantitative methods. A qualitative approach is used to describe the basic arithmetic mastery of students including numeracy, making Arithmetic / multiples numbers, complements especially nines and ten, and the concept of place values in numbers. The quantitative approach uses statistical tests with canonical correlation analysis to answer the relationship and influence between understanding of basic arithmetic mastery on the ability to solve math problems related to addition, subtraction, multiplication and division. From the results of the canonical weight and canonical loading function 1, it can be concluded that there is indeed a significant relationship between the dependent variate and the independent variate or basic arithmetic mastery and the students' ability to do math problems is indeed correlated together.


Sign in / Sign up

Export Citation Format

Share Document