Time Series Input Selection using Multiple Kernel Learning

In this paper we study the relevance of multiple kernel learning (MKL) for the automatic selection of time series inputs. Recently, MKL has gained great attention in the machine learning community due to its flexibility in modelling complex patterns and performing feature selection. In general, MKL constructs the kernel as a weighted linear combination of basis kernels, exploiting different sources of information. An efficient algorithm wrapping a Support Vector Regression model for optimizing the MKL weights, named SimpleMKL, is used for the analysis. In this sense, MKL performs feature selection by discarding inputs/kernels with low or null weights. The approach proposed is tested with simulated linear and nonlinear time series (AutoRegressive, Henon and Lorenz series).

Country portfolios and the Solow-model

This paper shows that an open economy Solow model provides a good description of international investment positions in industrialized countries. More than half of the variation of net foreign assets in the 1990's can be attributed to cross country differences in the savings rate, population and productivity growth. Furthermore, these factors seem to be an important channel through which output and wealth affect international investment positions. We interpret this funding as evidence that decreasing returns are an important source of international capital movements. The savings rate (andpopulation growth) influence the composition of country portfolios through their downward (upward) pressure on the marginal productivity of capital.

Unemployment in an estimated new Keynesian model

We reformulate the Smets-Wouters (2007) framework by embedding the theory of unemployment proposed in Galí (2011a,b). Weestimate the resulting model using postwar U.S. data, while treatingthe unemployment rate as an additional observable variable. Our approach overcomes the lack of identification of wage markup and laborsupply shocks highlighted by Chari, Kehoe and McGrattan (2008) intheir criticism of New Keynesian models, and allows us to estimate a"correct" measure of the output gap. In addition, the estimated modelcan be used to analyze the sources of unemployment fluctuations.

Real wage rigidities and the new Keynesian model

Most central banks perceive a trade-off between stabilizing inflation and stabilizing the gap between output and desired output. However, the standard new Keynesian framework implies no such trade-off. In that framework, stabilizing inflation is equivalent to stabilizing the welfare-relevant output gap. In this paper, we argue that this property of the new Keynesian framework, which we call the divine coincidence, is due to a special feature of the model: the absence of non trivial real imperfections.We focus on one such real imperfection, namely, real wage rigidities. When the baseline new Keynesian model is extended to allow for real wage rigidities, the divine coincidence disappears, and central banks indeed face a trade-off between stabilizing inflation and stabilizing the welfare-relevant output gap. We show that not only does the extended model have more realistic normative implications, but it also has appealing positive properties. In particular, it provides a natural interpretation for the dynamic inflation-unemployment relation found in the data.

A structural model of Australia as a small open economy

This paper sets up and estimates a structuralmodel of Australia as a small open economyusing Bayesian techniques. Unlike other recentstudies, the paper shows that a small microfoundedmodel can capture the open economydimensions quite well. Specifically, the modelattributes a substantial fraction of the volatilityof domestic output and inflation to foreigndisturbances, close to what is suggested by unrestrictedVAR studies. The paper also investigatesthe effects of various exogenous shockson the Australian economy.

Forecasting and turning point predictions in a Bayesian panel VAR model

We provide methods for forecasting variables and predicting turning points in panel Bayesian VARs. We specify a flexible model which accounts for both interdependencies in the cross section and time variations in the parameters. Posterior distributions for the parameters are obtained for a particular type of diffuse, for Minnesota-type and for hierarchical priors. Formulas for multistep, multiunit point and average forecasts are provided. An application to the problem of forecasting the growth rate of output and of predicting turning points in the G-7 illustrates the approach. A comparison with alternative forecasting methods is also provided.

A simple model of multiple equilibria based on risk

This paper shows how risk may aggravate fluctuations in economies with imperfect insurance and multiple assets. A two period job matching model is studied, in which risk averse agents act both as workers and as entrepreneurs. They choose between two types of investment: one type is riskless, while the other is a risky activity that creates jobs.Equilibrium is unique under full insurance. If investment is fully insured but unemployment risk is uninsured, then precautionary saving behavior dampens output fluctuations. However, if both investment and employment are uninsured, then an increase in unemployment gives agents an incentive to shift investment away from the risky asset, further increasing unemployment. This positive feedback may lead to multiple Pareto ranked equilibria. An overlapping generations version of the model may exhibit poverty traps or persistent multiplicity. Greater insurance is doubly beneficial in this context since it can both prevent multiplicity and promote risky investment.

Inflation and output dynamics with firm-owned capital

We model firm-owned capital in a stochastic dynamic New-Keynesian generalequilibrium model à la Calvo. We find that this structure impliesequilibrium dynamics which are quantitatively di¤erent from the onesassociated with a benchmark case where households accumulate capital andrent it to firms. Our findings therefore stress the importance ofmodeling an investment decision at the firm level in addition to ameaningful price setting decision. Along the way we argue that the problemof modeling firm-owned capital with Calvo price-setting has not been solvedin a correct way in the previous literature.

Extraction of soil water by plants: development and validation of a model

A quantitative model of water movement within the immediate vicinity of an individual root is developed and results of an experiment to validate the model are presented. The model is based on the assumption that the amount of water transpired by a plant in a certain period is replaced by an equal volume entering its root system during the same time. The model is based on the Darcy-Buckingham equation to calculate the soil water matric potential at any distance from a plant root as a function of parameters related to crop, soil and atmospheric conditions. The model output is compared against measurements of soil water depletion by rice roots monitored using γ-beam attenuation in a greenhouse of the Escola Superior de Agricultura "Luiz de Queiroz"/Universidade de São Paulo(ESALQ/USP) in Piracicaba, State of São Paulo, Brazil, in 1993. The experimental results are in agreement with the output from the model. Model simulations show that a single plant root is able to withdraw water from more than 0.1 m away within a few days. We therefore can assume that root distribution is a less important factor for soil water extraction efficiency.

Discordant increases in CD4+ T cells in human immunodeficiency virus-infected patients experiencing virologic treatment failure: role of changes in thymic output and T cell death.

Some patients infected with human immunodeficiency virus (HIV) who are experiencing antiretroviral treatment failure have persistent improvement in CD4+ T cell counts despite high plasma viremia. To explore the mechanisms responsible for this phenomenon, 2 parameters influencing the dynamics of CD4+ T cells were evaluated: death of mature CD4+ T cells and replenishment of the CD4+ T cell pool by the thymus. The improvement in CD4+ T cells observed in patients with treatment failure was not correlated with spontaneous, Fas ligand-induced, or activation-induced T cell death. In contrast, a significant correlation between the improvement in CD4+ T cell counts and thymic output, as assessed by measurement of T cell receptor excision circles, was observed. These observations suggest that increased thymic output contributes to the dissociation between CD4+ T cell counts and viremia in patients failing antiretroviral therapy and support a model in which drug-resistant HIV strains may have reduced replication rates and pathogenicity in the thymus.

Effect of expiratory flow limitation on respiratory mechanical impedance: a model study

Large phasic variations of respiratory mechanical impedance (Zrs) have been observed during induced expiratory flow limitation (EFL) (M. Vassiliou, R. Peslin, C. Saunier, and C. Duvivier. Eur. Respir. J. 9: 779-786, 1996). To clarify the meaning of Zrs during EFL, we have measured from 5 to 30 Hz the input impedance (Zin) of mechanical analogues of the respiratory system, including flow-limiting elements (FLE) made of easily collapsible rubber tubing. The pressures upstream (Pus) and downstream (Pds) from the FLE were controlled and systematically varied. Maximal flow (Vmax) increased linearly with Pus, was close to the value predicted from wave-speed theory, and was obtained for Pus-Pds of 4-6 hPa. The real part of Zin started increasing abruptly with flow (V) >85%Vmax and either further increased or suddenly decreased in the vicinity of V¿max. The imaginary part of Zin decreased markedly and suddenly above 95%Vmax. Similar variations of Zin during EFL were seen with an analogue that mimicked the changes of airway transmural pressure during breathing. After pressure andV measurements upstream and downstream from the FLE were combined, the latter was analyzed in terms of a serial (Zs) and a shunt (Zp) compartment. Zs was consistent with a large resistance and inertance, and Zp with a mainly elastic element having an elastance close to that of the tube walls. We conclude that Zrs data during EFL mainly reflect the properties of the FLE.

Essays on the measurement of school efficiency

Measuring school efficiency is a challenging task. First, a performance measurement technique has to be selected. Within Data Envelopment Analysis (DEA), one such technique, alternative models have been developed in order to deal with environmental variables. The majority of these models lead to diverging results. Second, the choice of input and output variables to be included in the efficiency analysis is often dictated by data availability. The choice of the variables remains an issue even when data is available. As a result, the choice of technique, model and variables is probably, and ultimately, a political judgement. Multi-criteria decision analysis methods can help the decision makers to select the most suitable model. The number of selection criteria should remain parsimonious and not be oriented towards the results of the models in order to avoid opportunistic behaviour. The selection criteria should also be backed by the literature or by an expert group. Once the most suitable model is identified, the principle of permanence of methods should be applied in order to avoid a change of practices over time. Within DEA, the two-stage model developed by Ray (1991) is the most convincing model which allows for an environmental adjustment. In this model, an efficiency analysis is conducted with DEA followed by an econometric analysis to explain the efficiency scores. An environmental variable of particular interest, tested in this thesis, consists of the fact that operations are held, for certain schools, on multiple sites. Results show that the fact of being located on more than one site has a negative influence on efficiency. A likely way to solve this negative influence would consist of improving the use of ICT in school management and teaching. Planning new schools should also consider the advantages of being located on a unique site, which allows reaching a critical size in terms of pupils and teachers. The fact that underprivileged pupils perform worse than privileged pupils has been public knowledge since Coleman et al. (1966). As a result, underprivileged pupils have a negative influence on school efficiency. This is confirmed by this thesis for the first time in Switzerland. Several countries have developed priority education policies in order to compensate for the negative impact of disadvantaged socioeconomic status on school performance. These policies have failed. As a result, other actions need to be taken. In order to define these actions, one has to identify the social-class differences which explain why disadvantaged children underperform. Childrearing and literary practices, health characteristics, housing stability and economic security influence pupil achievement. Rather than allocating more resources to schools, policymakers should therefore focus on related social policies. For instance, they could define pre-school, family, health, housing and benefits policies in order to improve the conditions for disadvantaged children.

Subclass problem-dependent design for error-correcting output codes

A common way to model multiclass classification problems is by means of Error-Correcting Output Codes (ECOCs). Given a multiclass problem, the ECOC technique designs a code word for each class, where each position of the code identifies the membership of the class for a given binary problem. A classification decision is obtained by assigning the label of the class with the closest code. One of the main requirements of the ECOC design is that the base classifier is capable of splitting each subgroup of classes from each binary problem. However, we cannot guarantee that a linear classifier model convex regions. Furthermore, nonlinear classifiers also fail to manage some type of surfaces. In this paper, we present a novel strategy to model multiclass classification problems using subclass information in the ECOC framework. Complex problems are solved by splitting the original set of classes into subclasses and embedding the binary problems in a problem-dependent ECOC design. Experimental results show that the proposed splitting procedure yields a better performance when the class overlap or the distribution of the training objects conceal the decision boundaries for the base classifier. The results are even more significant when one has a sufficiently large training size.

Visualization Resources for Iowa State University and the Iowa DOT: An Automated Design Model to Simulator Converter, November 2012

This project developed an automatic conversion software tool that takes input a from an Iowa Department of Transportation (DOT) MicroStation three-dimensional (3D) design file and converts it into a form that can be used by the University of Iowa’s National Advanced Driving Simulator (NADS) MiniSim. Once imported into the simulator, the new roadway has the identical geometric design features as in the Iowa DOT design file. The base roadway appears as a wireframe in the simulator software. Through additional software tools, textures and shading can be applied to the roadway surface and surrounding terrain to produce the visual appearance of an actual road. This tool enables Iowa DOT engineers to work with the universities to create drivable versions of prospective roadway designs. By driving the designs in the simulator, problems can be identified early in the design process. The simulated drives can also be used for public outreach and human factors driving research.

Simulation of growth and development of irrigated cowpea in Piauí State by CROPGRO model

The objective of this work was to adapt the CROPGRO model, which is part of the DSSAT system, for simulating the cowpea (Vigna unguiculata) growth and development under soil and climate conditions of the Baixo Parnaíba region, Piauí State, Brazil. In the CROPGRO, only input parameters that define crop species, cultivars, and ecotype were changed in order to characterize the cowpea crop. Soil and climate files were created for the considered site. Field experiments without water deficit were used to calibrate the model. In these experiments, dry matter (DM), leaf area index (LAI), yield components and grain yield of cowpea (cv. BR 14 Mulato) were evaluated. The results showed good fit for DM and LAI estimates. The medium values of R² and medium absolute error (MAE) were, respectively, 0.95 and 264.9 kg ha-1 for DM, and 0.97 and 0.22 for LAI. The difference between observed and simulated values of plant phenology varied from 0 to 3 days. The model also presented good performance for yield components simulation, excluding 100-grain weight, for which the error ranged from 20.9% to 34.3%. Considering the medium values of crop yield in two years, the model presented an error from 5.6%.