A mixture density network approach to modelling and exploiting uncertainty in nonlinear control problems

In nonlinear and stochastic control problems, learning an efficient feed-forward controller is not amenable to conventional neurocontrol methods. For these approaches, estimating and then incorporating uncertainty in the controller and feed-forward models can produce more robust control results. Here, we introduce a novel inversion-based neurocontroller for solving control problems involving uncertain nonlinear systems which could also compensate for multi-valued systems. The approach uses recent developments in neural networks, especially in the context of modelling statistical distributions, which are applied to forward and inverse plant models. Provided that certain conditions are met, an estimate of the intrinsic uncertainty for the outputs of neural networks can be obtained using the statistical properties of networks. More generally, multicomponent distributions can be modelled by the mixture density network. Based on importance sampling from these distributions a novel robust inverse control approach is obtained. This importance sampling provides a structured and principled approach to constrain the complexity of the search space for the ideal control law. The developed methodology circumvents the dynamic programming problem by using the predicted neural network uncertainty to localise the possible control solutions to consider. A nonlinear multi-variable system with different delays between the input-output pairs is used to demonstrate the successful application of the developed control algorithm. The proposed method is suitable for redundant control systems and allows us to model strongly non-Gaussian distributions of control signal as well as processes with hysteresis. © 2004 Elsevier Ltd. All rights reserved.

Viability Theory and Economic Modeling

A brief introduction into the theory of differential inclusions, viability theory and selections of set valued mappings is presented. As an application the implicit scheme of the Leontief dynamic input-output model is considered.

The Closed Loop Model by Regional Economics or How Big is the Ants’ Footprint?

The economy is communication between Man and Nature. It is an interaction-network between our outside and inside Nature, that is, the external Nature surrounding us and the internal nature expressing our human essence. Money is an institution of the society, an infrastructure that ensures division of labour, enables the flow of information and material between the participants. The concept of regional material and financial circular flow will be more important with the oncoming peak-oil and post-carbon era. We should describe in time the outlines of closed or semi-closed loops economy. The fundamentals of Input-Output will flourish once again; it could help us formulate the link between the efficiency and resiliency of a regional complex system.

Applying consumer responsibility principle in evaluating environmental load of carbon emissions

There is a need for a proper indicator in order to assess the environmental impact of international trade, therefore using the carbon footprint as an indicator can be relevant and useful. The aim of this study is to show from a methodological perspective how the carbon footprint, combined with input- output models can be used for analysing the impacts of international trade on the sustainable use of national resources in a country. The use of the input-output approach has the essential advantage of being able to track the transformation of goods through the economy. The study examines the environmental impact of consumption related to international trade, using the consumer responsibility principle. In this study the use of the carbon footprint and input-output methodology is shown on the example of the Hungarian consumption and the impact of international trade. Moving from a production- based approach in climate policy to a consumption-perspective principle and allocation, would also help to increase the efficiency of emission reduction targets and the evaluation of the ecological impacts of international trade.

A nemzetközi kereskedelem környezeti hatásainak vizsgálata a karbon lábnyom segítségével

Napjainkban már köztudott tény az, hogy az antropogén szén-dioxid kibocsátás nagymértékben hozzájárul a klímaváltozáshoz. Ahhoz, hogy ténylegesen értékelni tudjuk a kibocsátás környezeti hatását, szükség van egy olyan indikátorra, ami figyelembe veszi a természeti erőforrások és a természet megújuló-képessége által szabott korlátokat. A karbon lábnyom, ilyen módon egy releváns indikátor erre, és segítségével kimutatható, hogy mennyire felel meg egy ország termelése, életmódja, fogyasztási szerkezete a fenntarthatóság kritériumainak. Ennek a tanulmánynak a célja, hogy megvizsgálja a magyarországi fogyasztás környezetterhelésének szerkezetét a karbon lábnyom indikátorának segítségével, középpontba helyezve a nemzetközi kereskedelem hazai hatásainak vizsgálatát. A nemzetközi kereskedelemben elfoglalt pozíció, nemcsak egy ország gazdasági szerkezetét és versenyképességét határozza meg, hanem erőteljes hatással van a fogyasztói szokásokra és a fogyasztás környezetterhelésére is. A karbon lábnyom tartalmazza annak a környezetterhelésnek az értékeit is, amely az importált termékek és szolgáltatások elfogyasztásából származik. Fontos megvizsgálni azt, hogy mely szektorokban van nagy jelentősége az importált termékek környezetterhelésének és ez hogyan járul hozzá a hazai fogyasztási mintákhoz. A tanulmány a fogyasztás környezetterhelését a fogyasztói felelősségi elvet alkalmazva vizsgálja. A karbon lábnyom fogyasztói szempontból való vizsgálata azért jelentős, mert ezáltal képes felhívni a döntéshozók figyelmét, hogy mely szektorokban és fogyasztási kategóriákban jelenik meg nem fenntartható fogyasztás. Így, felszínre kerülnek azok a területek, ahol erőteljes beavatkozásra van szükség a kibocsátások csökkentése érdekében, mindez nemcsak a termelői oldal, hanem a fogyasztói magatartás befolyásolása, jobb informálása, oktatása által. A tanulmány módszertanában kombináltan alkalmazza az ökológiai lábnyom számításból származó karbon lábnyom számítás módszertanát, kiegészítve az ágazati kapcsolatok mérlegének (input-output táblázatok) módszertanával. Ez a kombinált módszertan, a szoros ágazati összefüggéseket és kölcsönhatásokat is kezelni tudja, megjelenítve mind a direkt és indirekt környezeti hatásokat is.

Applying consumer responsibility principle in evaluating environmental load of carbon emissions

There is a need for a proper indicator in order to assess the environmental impact of international trade, therefore using the carbon footprint as an indicator can be relevant and useful. The aim of this study is to show from a methodological perspective how the carbon footprint, combined with input- output models can be used for analysing the impacts of international trade on the sustainable use of national resources in a country. The use of the input-output approach has the essential advantage of being able to track the transformation of goods through the economy. The study examines the environmental impact of consumption related to international trade, using the consumer responsibility principle. In this study the use of the carbon footprint and input-output methodology is shown on the example of the Hungarian consumption and the impact of international trade. Moving from a production- based approach in climate policy to a consumption-perspective principle and allocation, would also help to increase the efficiency of emission reduction targets and the evaluation of the ecological impacts of international trade.

Developing a high density platinum/alumina hermetic feedthrough

Typically, hermetic feedthroughs for implantable devices, such as pacemakers, use a alumina ceramic insulator brazed to a platinum wire pin. This combination of material has a long history in implantable devices and has been approved by the FDA for implantable hermetic feedthroughs. The growing demand for increased input/output (I/O) hermetic feedthroughs for implantable neural stimulator applications could be addressed by developing a new, cofired platinum/alumina multilayer ceramic technology in a configuration that supports 300 plus I/Os, which is not commercially available. Seven platinum powders with different particle sizes were used to develop different conductive cofire inks to control the densification mismatch between platinum and alumina. Firing profile (ramp rate, burn- out and holding times) and firing atmosphere and concentrations (hydrogen (wet/dry), air, neutral, vacuum) were also optimized. Platinum and alumina exhibit the alloy formation reaction in a reduced atmosphere. Formation of any compound can increase the bonding of the metal/ceramic interface, resulting in enhanced hermeticity. The feedthrough fabricated in a reduced atmosphere demonstrated significantly superior performance than that of other atmospheres. A composite structure of tungsten/platinum ratios graded thru the via structure (pure W, 50/50 W/Pt, 80/20 Pt/W and pure Pt) exhibited the best performance in comparison to the performance of other materials used for ink metallization. Studies on the high temperature reaction of platinum and alumina, previously unreported, showed that, at low temperatures in reduced atmosphere, Pt 3Al or Pt8Al21 with a tetragonal structure would be formed. Cubic Pt3Al is formed upon heating the sample to temperatures above 1350 °C. This cubic structure is the equilibrium state of Pt-Al alloy at high temperatures. The alumina dissolves into the platinum ink and is redeposited as a surface coating. This was observed on both cofired samples and pure platinum thin films coated on a 99.6 Wt% alumina and fired at 1550 °C. Different mechanisms are proposed to describe this behavior based on the size of the platinum particle

Produção de narrativas escritas no ensino fundamental: estudo de casos

This paper focuses on the construction of the narrative for elementary school students, searching to identify strategies they employed in the production of narrative texts representative of miniconto genre. Therefore, we take a sample forty texts produced by students of 6 and 9 years of basic education, twenty in the 6th year students (ten public school and ten private school) and twenty students in 9th grade (distributed similarly between public education and private). In general, we aim to understand the mechanisms by which producers build their narratives, as well as providing input for analysis of textual production of this genre. This research is based on Functional-Linguistic assumptions of the American side, inspired by Givón (2001), Thompson (2005), Hopper (1987), Bybee (2010), Traugott (2003), Martelotta (2008), Furtado da Cunha (2011), among others. In addition, from the theoretical framework presented by Labov (1972) about the narrative, coupled with Batoréo contribution (1998), we observed the recurring elements in the structure of narratives under study: abstract, orientation, complication, resolution, evaluation and coda. Also approached, but that in a complementary way, the notion of gender presented in Marcuschi (2002). This is a research quantitative and qualitative, with descriptive and analytical-interpretive bias. In corpus analysis, we consider the following categories: gender discourse miniconto; compositional structure of the narrative; informativeness (discursive progression, thematic coherence and narrative, topical-referential distribution); informative relevance (figure / ground). At the end of the work, our initial hypothesis of the better performance of students in 9th grade, compared to 6, and the particular context of education in relation to the public context, not confirmed, since, in the comparative study revealed that the groups have similar performance as the construction of the narrative, making use of the same strategies in its construction.

Produção de narrativas escritas no ensino fundamental: estudo de casos

This paper focuses on the construction of the narrative for elementary school students, searching to identify strategies they employed in the production of narrative texts representative of miniconto genre. Therefore, we take a sample forty texts produced by students of 6 and 9 years of basic education, twenty in the 6th year students (ten public school and ten private school) and twenty students in 9th grade (distributed similarly between public education and private). In general, we aim to understand the mechanisms by which producers build their narratives, as well as providing input for analysis of textual production of this genre. This research is based on Functional-Linguistic assumptions of the American side, inspired by Givón (2001), Thompson (2005), Hopper (1987), Bybee (2010), Traugott (2003), Martelotta (2008), Furtado da Cunha (2011), among others. In addition, from the theoretical framework presented by Labov (1972) about the narrative, coupled with Batoréo contribution (1998), we observed the recurring elements in the structure of narratives under study: abstract, orientation, complication, resolution, evaluation and coda. Also approached, but that in a complementary way, the notion of gender presented in Marcuschi (2002). This is a research quantitative and qualitative, with descriptive and analytical-interpretive bias. In corpus analysis, we consider the following categories: gender discourse miniconto; compositional structure of the narrative; informativeness (discursive progression, thematic coherence and narrative, topical-referential distribution); informative relevance (figure / ground). At the end of the work, our initial hypothesis of the better performance of students in 9th grade, compared to 6, and the particular context of education in relation to the public context, not confirmed, since, in the comparative study revealed that the groups have similar performance as the construction of the narrative, making use of the same strategies in its construction.

Interaction dynamics in small networks of nonlinear elements

We study a small circuit of coupled nonlinear elements to investigate general features of signal transmission through networks. The small circuit itself is perceived as building block for larger networks. Individual dynamics and coupling are motivated by neuronal systems: We consider two types of dynamical modes for an individual element, regular spiking and chattering and each individual element can receive excitatory and/or inhibitory inputs and is subjected to different feedback types (excitatory and inhibitory; forward and recurrent). Both, deterministic and stochastic simulations are carried out to study the input-output relationships of these networks. Major results for regular spiking elements include frequency locking, spike rate amplification for strong synaptic coupling, and inhibition-induced spike rate control which can be interpreted as a output frequency rectification. For chattering elements, spike rate amplification for low frequencies and silencing for large frequencies is characteristic

Distributed Extremum-Seeking Control with Applications to High-Altitude Balloons

The real-time optimization of large-scale systems is a difficult problem due to the need for complex models involving uncertain parameters and the high computational cost of solving such problems by a decentralized approach. Extremum-seeking control (ESC) is a model-free real-time optimization technique which can estimate unknown parameters and can optimize nonlinear time-varying systems using only a measurement of the cost function to be minimized. In this thesis, we develop a distributed version of extremum-seeking control which allows large-scale systems to be optimized without models and with minimal computing power. First, we develop a continuous-time distributed extremum-seeking controller. It has three main components: consensus, parameter estimation, and optimization. The consensus provides each local controller with an estimate of the cost to be minimized, allowing them to coordinate their actions. Using this cost estimate, parameters for a local input-output model are estimated, and the cost is minimized by following a gradient descent based on the estimate of the gradient. Next, a similar distributed extremum-seeking controller is developed in discrete-time. Finally, we consider an interesting application of distributed ESC: formation control of high-altitude balloons for high-speed wireless internet. These balloons must be steered into a favourable formation where they are spread out over the Earth and provide coverage to the entire planet. Distributed ESC is applied to this problem, and is shown to be effective for a system of 1200 ballons subjected to realistic wind currents. The approach does not require a wind model and uses a cost function based on a Voronoi partition of the sphere. Distributed ESC is able to steer balloons from a few initial launch sites into a formation which provides coverage to the entire Earth and can maintain a similar formation as the balloons move with the wind around the Earth.

New Developments in Tourism and Hotel Demand Modeling and Forecasting

Abstract Purpose The purpose of the study is to review recent studies published from 2007-2015 on tourism and hotel demand modeling and forecasting with a view to identifying the emerging topics and methods studied and to pointing future research directions in the field. Design/Methodology/approach Articles on tourism and hotel demand modeling and forecasting published in both science citation index (SCI) and social science citation index (SSCI) journals were identified and analyzed. Findings This review found that the studies focused on hotel demand are relatively less than those on tourism demand. It is also observed that more and more studies have moved away from the aggregate tourism demand analysis, while disaggregate markets and niche products have attracted increasing attention. Some studies have gone beyond neoclassical economic theory to seek additional explanations of the dynamics of tourism and hotel demand, such as environmental factors, tourist online behavior and consumer confidence indicators, among others. More sophisticated techniques such as nonlinear smooth transition regression, mixed-frequency modeling technique and nonparametric singular spectrum analysis have also been introduced to this research area. Research limitations/implications The main limitation of this review is that the articles included in this study only cover the English literature. Future review of this kind should also include articles published in other languages. The review provides a useful guide for researchers who are interested in future research on tourism and hotel demand modeling and forecasting. Practical implications This review provides important suggestions and recommendations for improving the efficiency of tourism and hospitality management practices. Originality/value The value of this review is that it identifies the current trends in tourism and hotel demand modeling and forecasting research and points out future research directions.

Use of a Neural Network to Predict Strength and Optimum Compositions of Natural Alumina-Silica-Based Geopolymers

In order to predict compressive strength of geopolymers prepared from alumina-silica natural products, based on the effect of Al 2 O 3 /SiO 2, Na 2 O/Al 2 O 3, Na 2 O/H 2 O, and Na/[Na+K], more than 50 pieces of data were gathered from the literature. The data was utilized to train and test a multilayer artificial neural network (ANN). Therefore a multilayer feedforward network was designed with chemical compositions of alumina silicate and alkali activators as inputs and compressive strength as output. In this study, a feedforward network with various numbers of hidden layers and neurons were tested to select the optimum network architecture. The developed three-layer neural network simulator model used the feedforward back propagation architecture, demonstrated its ability in training the given input/output patterns. The cross-validation data was used to show the validity and high prediction accuracy of the network. This leads to the optimum chemical composition and the best paste can be made from activated alumina-silica natural products using alkaline hydroxide, and alkaline silicate. The research results are in agreement with mechanism of geopolymerization.


Read More: http://ascelibrary.org/doi/abs/10.1061/(ASCE)MT.1943-5533.0000829

Alcohol-Related Problems in Eastern Europe : A Comparative Perspective

This thesis investigates the association between alcohol consumption and alcohol-related harm in Eastern Europe. The main aim was to estimate to what extent changes in per capita alcohol consumption have an impact on different forms of alcohol-related mortality, and to put the results in an international comparative perspective. The thesis includes four papers; the first two papers use aggregate time-series analysis to assess how changes in per capita consumption affect rates in suicide mortality and fatal non-intentional injuries in several Eastern European countries, respectively. The third paper applies the same methodological approach to analyse the population-level relationship between alcohol and homicide in Russia and the U.S.. The fourth paper employs survey data to assess how the risk of experiencing alcohol-related problems in relation to volume of consumption in the Baltic countries compares to Sweden and Italy. The results of the first three papers suggests: (i) that changes in per capita consumption are significantly related to changes in mortality rates of suicide, non-intentional injuries and homicide in the countries under study; (ii) that the relationship is stronger for men than for women, and (iii) that the relationship tends to be stronger in the countries with more detrimental drinking patterns, e.g. Russia. The results of the fourth paper suggest that the risk of experiencing alcohol-related problems in relation to level of drinking in the Baltic countries is similar to the corresponding risk in Sweden, but considerably stronger than in Italy. In conclusion, the findings support the significance of a public health approach to alcohol-related problems in Eastern Europe, i.e., policy measures directed towards total alcohol consumption. In addition, strategies aimed at reducing the occurrence of binge drinking seem to have great potential for reducing alcohol-related harm and mortality in Eastern European countries.

SEQUENTIAL DECISIONS AND PREDICTIONS IN NATURAL LANGUAGE PROCESSING

Natural language processing has achieved great success in a wide range of ap- plications, producing both commercial language services and open-source language tools. However, most methods take a static or batch approach, assuming that the model has all information it needs and makes a one-time prediction. In this disser- tation, we study dynamic problems where the input comes in a sequence instead of all at once, and the output must be produced while the input is arriving. In these problems, predictions are often made based only on partial information. We see this dynamic setting in many real-time, interactive applications. These problems usually involve a trade-off between the amount of input received (cost) and the quality of the output prediction (accuracy). Therefore, the evaluation considers both objectives (e.g., plotting a Pareto curve). Our goal is to develop a formal understanding of sequential prediction and decision-making problems in natural language processing and to propose efficient solutions. Toward this end, we present meta-algorithms that take an existent batch model and produce a dynamic model to handle sequential inputs and outputs. Webuild our framework upon theories of Markov Decision Process (MDP), which allows learning to trade off competing objectives in a principled way. The main machine learning techniques we use are from imitation learning and reinforcement learning, and we advance current techniques to tackle problems arising in our settings. We evaluate our algorithm on a variety of applications, including dependency parsing, machine translation, and question answering. We show that our approach achieves a better cost-accuracy trade-off than the batch approach and heuristic-based decision- making approaches. We first propose a general framework for cost-sensitive prediction, where dif- ferent parts of the input come at different costs. We formulate a decision-making process that selects pieces of the input sequentially, and the selection is adaptive to each instance. Our approach is evaluated on both standard classification tasks and a structured prediction task (dependency parsing). We show that it achieves similar prediction quality to methods that use all input, while inducing a much smaller cost. Next, we extend the framework to problems where the input is revealed incremen- tally in a fixed order. We study two applications: simultaneous machine translation and quiz bowl (incremental text classification). We discuss challenges in this set- ting and show that adding domain knowledge eases the decision-making problem. A central theme throughout the chapters is an MDP formulation of a challenging problem with sequential input/output and trade-off decisions, accompanied by a learning algorithm that solves the MDP.