Invitation Only Research Event

Event participants

Richard King, Senior Research Fellow, Energy, Environment and Resources Department, Chatham House
Chair: Duncan Brack, Associate Fellow, Energy, Environment and Resources Department, Chatham House

29 January 2020

Policymakers are in danger of sleepwalking into ineffective carbon dioxide removal solutions in the quest to tackle climate change. This paper warns against overreliance on bioenergy with carbon capture and storage (BECCS). 

Danilo Covaes Nogarotto, Caio Lucidius Naberezny Azevedo, Jorge Luis Bazán.

Source: Brazilian Journal of Probability and Statistics, Volume 34, Number 2, 304--322.

Abstract:
The interest on the analysis of the zero–one augmented beta regression (ZOABR) model has been increasing over the last few years. In this work, we developed a Bayesian inference for the ZOABR model, providing some contributions, namely: we explored the use of Jeffreys-rule and independence Jeffreys prior for some of the parameters, performing a sensitivity study of prior choice, comparing the Bayesian estimates with the maximum likelihood ones and measuring the accuracy of the estimates under several scenarios of interest. The results indicate, in a general way, that: the Bayesian approach, under the Jeffreys-rule prior, was as accurate as the ML one. Also, different from other approaches, we use the predictive distribution of the response to implement Bayesian residuals. To further illustrate the advantages of our approach, we conduct an analysis of a real psychometric data set including a Bayesian residual analysis, where it is shown that misleading inference can be obtained when the data is transformed. That is, when the zeros and ones are transformed to suitable values and the usual beta regression model is considered, instead of the ZOABR model. Finally, future developments are discussed.

Weiping Zhang, Jiangli Wang, Fang Qian, Yu Chen.

Source: Brazilian Journal of Probability and Statistics, Volume 34, Number 1, 35--50.

Abstract:
Longitudinal zero-inflated count data are widely encountered in many fields, while modeling the correlation between measurements for the same subject is more challenge due to the lack of suitable multivariate joint distributions. This paper studies a novel mean-correlation modeling approach for longitudinal zero-inflated regression model, solving both problems of specifying joint distribution and parsimoniously modeling correlations with no constraint. The joint distribution of zero-inflated discrete longitudinal responses is modeled by a copula model whose correlation parameters are innovatively represented in hyper-spherical coordinates. To overcome the computational intractability in maximizing the full likelihood function of the model, we further propose a computationally efficient pairwise likelihood approach. We then propose separated mean and correlation regression models to model these key quantities, such modeling approach can also handle irregularly and possibly subject-specific times points. The resulting estimators are shown to be consistent and asymptotically normal. Data example and simulations support the effectiveness of the proposed approach.

Wesley Bertoli, Katiane S. Conceição, Marinho G. Andrade, Francisco Louzada.

Source: Brazilian Journal of Probability and Statistics, Volume 33, Number 4, 826--860.

Abstract:
The primary goal of this paper is to introduce the zero-modified Poisson–Lindley regression model as an alternative to model overdispersed count data exhibiting inflation or deflation of zeros in the presence of covariates. The zero-modification is incorporated by considering that a zero-truncated process produces positive observations and consequently, the proposed model can be fitted without any previous information about the zero-modification present in a given dataset. A fully Bayesian approach based on the g-prior method has been considered for inference concerns. An intensive Monte Carlo simulation study has been conducted to evaluate the performance of the developed methodology and the maximum likelihood estimators. The proposed model was considered for the analysis of a real dataset on the number of bids received by $126$ U.S. firms between 1978–1985, and the impact of choosing different prior distributions for the regression coefficients has been studied. A sensitivity analysis to detect influential points has been performed based on the Kullback–Leibler divergence. A general comparison with some well-known regression models for discrete data has been presented.

Model Stealing (MS) attacks allow an adversary with black-box access to a Machine Learning model to replicate its functionality, compromising the confidentiality of the model. Such attacks train a clone model by using the predictions of the target model for different inputs. The effectiveness of such attacks relies heavily on the availability of data necessary to query the target model. Existing attacks either assume partial access to the dataset of the target model or availability of an alternate dataset with semantic similarities.

This paper proposes MAZE -- a data-free model stealing attack using zeroth-order gradient estimation. In contrast to prior works, MAZE does not require any data and instead creates synthetic data using a generative model. Inspired by recent works in data-free Knowledge Distillation (KD), we train the generative model using a disagreement objective to produce inputs that maximize disagreement between the clone and the target model. However, unlike the white-box setting of KD, where the gradient information is available, training a generator for model stealing requires performing black-box optimization, as it involves accessing the target model under attack. MAZE relies on zeroth-order gradient estimation to perform this optimization and enables a highly accurate MS attack.

Our evaluation with four datasets shows that MAZE provides a normalized clone accuracy in the range of 0.91x to 0.99x, and outperforms even the recent attacks that rely on partial data (JBDA, clone accuracy 0.13x to 0.69x) and surrogate data (KnockoffNets, clone accuracy 0.52x to 0.97x). We also study an extension of MAZE in the partial-data setting and develop MAZE-PD, which generates synthetic data closer to the target distribution. MAZE-PD further improves the clone accuracy (0.97x to 1.0x) and reduces the query required for the attack by 2x-24x.

Boyu Ren, Sergio Bacallado, Stefano Favaro, Tommi Vatanen, Curtis Huttenhower, Lorenzo Trippa.

Source: The Annals of Applied Statistics, Volume 14, Number 1, 494--517.

Abstract:
Detecting associations between microbial compositions and sample characteristics is one of the most important tasks in microbiome studies. Most of the existing methods apply univariate models to single microbial species separately, with adjustments for multiple hypothesis testing. We propose a Bayesian analysis for a generalized mixed effects linear model tailored to this application. The marginal prior on each microbial composition is a Dirichlet process, and dependence across compositions is induced through a linear combination of individual covariates, such as disease biomarkers or the subject’s age, and latent factors. The latent factors capture residual variability and their dimensionality is learned from the data in a fully Bayesian procedure. The proposed model is tested in data analyses and simulation studies with zero-inflated compositions. In these settings and within each sample, a large proportion of counts per microbial species are equal to zero. In our Bayesian model a priori the probability of compositions with absent microbial species is strictly positive. We propose an efficient algorithm to sample from the posterior and visualizations of model parameters which reveal associations between covariates and microbial compositions. We evaluate the proposed method in simulation studies, and then analyze a microbiome dataset for infants with type 1 diabetes which contains a large proportion of zeros in the sample-specific microbial compositions.

Brian Neelon.

Source: Bayesian Analysis, Volume 14, Number 3, 849--875.

Abstract:
Motivated by a study examining spatiotemporal patterns in inpatient hospitalizations, we propose an efficient Bayesian approach for fitting zero-inflated negative binomial models. To facilitate posterior sampling, we introduce a set of latent variables that are represented as scale mixtures of normals, where the precision terms follow independent Pólya-Gamma distributions. Conditional on the latent variables, inference proceeds via straightforward Gibbs sampling. For fixed-effects models, our approach is comparable to existing methods. However, our model can accommodate more complex data structures, including multivariate and spatiotemporal data, settings in which current approaches often fail due to computational challenges. Using simulation studies, we highlight key features of the method and compare its performance to other estimation procedures. We apply the approach to a spatiotemporal analysis examining the number of annual inpatient admissions among United States veterans with type 2 diabetes.

Lei Liu, Ya-Chen Tina Shih, Robert L. Strawderman, Daowen Zhang, Bankole A. Johnson, Haitao Chai.

Source: Statistical Science, Volume 34, Number 2, 253--279.

Abstract:
Zero-inflated nonnegative continuous (or semicontinuous) data arise frequently in biomedical, economical, and ecological studies. Examples include substance abuse, medical costs, medical care utilization, biomarkers (e.g., CD4 cell counts, coronary artery calcium scores), single cell gene expression rates, and (relative) abundance of microbiome. Such data are often characterized by the presence of a large portion of zero values and positive continuous values that are skewed to the right and heteroscedastic. Both of these features suggest that no simple parametric distribution may be suitable for modeling such type of outcomes. In this paper, we review statistical methods for analyzing zero-inflated nonnegative outcome data. We will start with the cross-sectional setting, discussing ways to separate zero and positive values and introducing flexible models to characterize right skewness and heteroscedasticity in the positive values. We will then present models of correlated zero-inflated nonnegative continuous data, using random effects to tackle the correlation on repeated measures from the same subject and that across different parts of the model. We will also discuss expansion to related topics, for example, zero-inflated count and survival data, nonlinear covariate effects, and joint models of longitudinal zero-inflated nonnegative continuous data and survival. Finally, we will present applications to three real datasets (i.e., microbiome, medical costs, and alcohol drinking) to illustrate these methods. Example code will be provided to facilitate applications of these methods.

FAO Director-General José Graziano da Silva underlined his firm belief that a hunger-free world is possible "within our lifetimes," during high-level talks in New York. "The Zero Hunger Challenge calls for something new – something bold, but long overdue," he said. It was a "decisive global commitment to end hunger, eliminate childhood stunting, make all food systems sustainable, eradicate rural poverty, [...]

FAO’s emergency programme to support farmers affected by severe flooding in northern Benin includes creating resilience among communities in order to avert further threats to their livelihoods and food security. Farming families in northern Benin lost crops, livestock and fishing grounds when the Niger River overran its banks in August, just as many villagers were trying to recover from previous floods [...]

The mission for an end to hunger in the world’s most populous region has received a boost, with member countries responding positively to a call by FAO for a “massive effort” to end hunger in Asia and the Pacific. 1. Asia-Pacific is home to nearly two-thirds of the world’s chronically hungry people. |True|     Asia-Pacific, with over 4.2 billion people, is home [...]

At the Rio+20 Conference on Sustainable Development in June 2012, UN Secretary-General Ban Ki-moon announced a new global challenge for world leaders and individuals from all sectors: create a world where no one is hungry. He emphasized that there is enough food in the world to feed our population, so the challenge comes from making sure that everyone has access [...]

Have you ever thought about producing a video on food, nutrition, sustainability or hunger? Whether you’re a food buff, a student, an activist, movie geek or professional filmmaker, we have just the thing for you.Short Food Movie is a global open call for videos inspired by the theme for Expo Milano 2015, “Feeding the Planet. Energy for Life.”  It includes [...]

Radio culture is gaining more and more ground as millions of listeners take to audio podcasts as a convenient and accessible way to learn new information.  Which is why FAO is stepping up into the medium to bring you insights into some of the issues concerning food and agriculture worldwide. Here are the first seven audio offerings of FAO’s new podcast [...]

Food waste has become a dangerous habit: buying more than we need at supermarkets, letting fruits and vegetables spoil at home or ordering more than we can eat at restaurants.  Each year, about 1/3 of the food we produce globally is lost or wasted. In developing countries, a large part of this food (40%) is lost at the harvest or processing [...]

Constituting only 5 percent of the world population, indigenous peoples nevertheless are vital stewards of the environment. Traditional indigenous territories encompass 22 percent of the world’s land surface, but 80 percent of the planet’s biodiversity.  A third of global forests, crucial for curbing gas emissions, are primarily managed by indigenous peoples, families, smallholders and local communities. Indigenous foods are also particularly [...]

Perhaps from outer space, it is easy to get the perspective that we only have one earth and that it is succumbing to climate change. Seeing the earth from space though is a feat that, unfortunately, most of us will never accomplish. We have to rely on Astronaut Thomas Pesquet and the other brave women and men astronauts to provide [...]

A new breed of food establishment is attempting to do away with food waste entirely

 Prudential norms on Investment in Zero Coupon Bonds

NEW CASTLE (Jan. 30, 2019) With Delaware facing the coldest temperatures of the winter from Wednesday night through Thursday, Code Purple is being declared across the state by the nonprofit and volunteer organizations that operate the shelters to keep people who are homeless safe during dangerous winter weather. When severe snowstorms or sub-zero temperatures are […]

Electric truck by ABB showcased in Swiss capital, Bern.

According to the Pradhan Mantri Jan Dhan Yojana (PMJDY) website, the total number of RuPay cards issued so far stands at over 290 million.

A fully Keynesian approach will need to be taken by the government. People are losing jobs and production activity has come to a standstill due to the lockdown.

Is there a better approach to varying the coefficients of a digital IIR over time to adjust the values of its poles and zeros than just recalculating the whole thing every time it changes? For example, lots of synth programs can apply an LFO to the cutoff frequency of a low/high pass filter. I can do some polynomial multiplication to get the coefficients for an IIR filter given its poles and zeros, but am wondering if there is a better way to adjust them over time than simply doing all the calculations over again for new poles/zeros. Particularly, I'm curious if there is a method that will more or less work for an arbitrary number of poles and zeros. You could use a filter implementation (state space) that directly uses the pole/zero values instead of a polynomial walmartone. That might be computationally more expensive, though (as you are taking a trip through the domain of complex numbers even though your inputs and output are real), and possibly numerically iffy.As far as I am aware, modifying filter behavior while introducing as few artefacts as possible is still an area of research. You might get away with just adjusting the filter coefficients if you do it slowly, but this does not mean this is the best method.In an audio application, I assume they do not switch filter coefficients abruptly, but instead do a cross-fade between the (settled) first filter and the (mostly or completely settled) target filter to avoid audible artefacts.

●Electric air mobility pioneer used the 3DEXPERIENCE platform on the cloud to develop prototype in two years ●“Reinvent the Sky” industry solution experience provides full data security in a single, standards-based environment ●Dassault Systèmes enables companies of all sizes to create new categories of sustainable air mobility systems that will change how the world travels

The ZeroNights 2019 Call For Papers has been announced. It will be held in Saint-Petersburg, Russia November 12th through the 13th, 2019.

Zero Day Initiative Advisory 10-231 - This vulnerability allows remote attackers to execute arbitrary code on vulnerable installations of Juniper SA Series devices. Authentication is not required to exploit this vulnerability. The specific flaw exists within the meeting_testjava.cgi page which is used to test JVM compatibility. When handling the DSID HTTP header the code allows an attacker to inject arbitrary javascript into the page. This can be abused by an attacker to perform a cross-site scripting attack on the device.