Semi-supervised learning of hierarchical latent trait models for data visualisation

An interactive hierarchical Generative Topographic Mapping (HGTM) ¸iteH_GTM has been developed to visualise complex data sets. In this paper, we build a more general visualisation system by extending the HGTM visualisation system in 3 directions: bf (1) We generalize HGTM to noise models from the exponential family of distributions. The basic building block is the Latent Trait Model (LTM) developed in ¸iteKaban_pami. bf (2) We give the user a choice of initializing the child plots of the current plot in either em interactive, or em automatic mode. In the interactive mode the user interactively selects ``regions of interest'' as in ¸iteH_GTM, whereas in the automatic mode an unsupervised minimum message length (MML)-driven construction of a mixture of LTMs is employed. bf (3) We derive general formulas for magnification factors in latent trait models. Magnification factors are a useful tool to improve our understanding of the visualisation plots, since they can highlight the boundaries between data clusters. The unsupervised construction is particularly useful when high-level plots are covered with dense clusters of highly overlapping data projections, making it difficult to use the interactive mode. Such a situation often arises when visualizing large data sets. We illustrate our approach on a toy example and apply our system to three more complex real data sets.

Semisupervised learning of hierarchical latent trait models for data visualization

Recently, we have developed the hierarchical Generative Topographic Mapping (HGTM), an interactive method for visualization of large high-dimensional real-valued data sets. In this paper, we propose a more general visualization system by extending HGTM in three ways, which allows the user to visualize a wider range of data sets and better support the model development process. 1) We integrate HGTM with noise models from the exponential family of distributions. The basic building block is the Latent Trait Model (LTM). This enables us to visualize data of inherently discrete nature, e.g., collections of documents, in a hierarchical manner. 2) We give the user a choice of initializing the child plots of the current plot in either interactive, or automatic mode. In the interactive mode, the user selects "regions of interest," whereas in the automatic mode, an unsupervised minimum message length (MML)-inspired construction of a mixture of LTMs is employed. The unsupervised construction is particularly useful when high-level plots are covered with dense clusters of highly overlapping data projections, making it difficult to use the interactive mode. Such a situation often arises when visualizing large data sets. 3) We derive general formulas for magnification factors in latent trait models. Magnification factors are a useful tool to improve our understanding of the visualization plots, since they can highlight the boundaries between data clusters. We illustrate our approach on a toy example and evaluate it on three more complex real data sets. © 2005 IEEE.

Robust Parametric Estimation of Branching Processes with a Random Number of Ancestors

2000 Mathematics Subject Classification: 60J80.

Parametric Estimation in Branching Processes with an Increasing Random Number of Ancestors

2000 Mathematics Subject Classification: 60J80, 62M05

Techno-economic analysis of three HVAC retrofitting options

Accounting for around 40% of the total final energy consumption, the building stock is an important area of focus on the way to reaching the energy goals set for the European Union. The relatively small share of new buildings makes renovation of existing buildings possibly the most feasible way of improving the overall energy performance of the building stock. This of course involves improvements on the climate shell, for example by additional insulation or change of window glazing, but also installation of new heating systems, to increase the energy efficiency and to fit the new heat load after renovation. In the choice of systems for heating, ventilation and air conditioning (HVAC), it is important to consider their performance for space heating as well as for domestic hot water (DHW), especially for a renovated house where the DHW share of the total heating consumption is larger. The present study treats the retrofitting of a generic single family house, which was defined as a reference building in a European energy renovation project. Three HVAC retrofitting options were compared from a techno-economic point of view: A) Air-to-water heat pump (AWHP) and mechanical ventilation with heat recovery (MVHR), B) Exhaust air heat pump (EAHP) with low-temperature ventilation radiators, and C) Gas boiler and ventilation with MVHR. The systems were simulated for houses with two levels of heating demand and four different locations: Stockholm, Gdansk, Stuttgart and London. They were then evaluated by means of life cycle cost (LCC) and primary energy consumption. Dynamic simulations were done in TRNSYS 17. In most cases, system C with gas boiler and MVHR was found to be the cheapest retrofitting option from a life cycle perspective. The advantage over the heat pump systems was particularly clear for a house in Germany, due to the large discrepancy between national prices of natural gas and electricity. In Sweden, where the price difference is much smaller, the heat pump systems had almost as low or even lower life cycle costs than the gas boiler system. Considering the limited availability of natural gas in Sweden, systems A and B would be the better options. From a primary energy point of view system A was the best option throughout, while system B often had the highest primary energy consumption. The limited capacity of the EAHP forced it to use more auxiliary heating than the other systems did, which lowered its COP. The AWHP managed the DHW load better due to a higher capacity, but had a lower COP than the EAHP in space heating mode. Systems A and C were notably favoured by the air heat recovery, which significantly reduced the heating demand. It was also seen that the DHW share of the total heating consumption was, as expected, larger for the house with the lower space heating demand. This confirms the supposition that it is important to include DHW in the study of HVAC systems for retrofitting.

A family of autoregressive conditional duration models

This paper develops a family of autoregressive conditional duration (ACD) models that encompasses most specifications in the literature. The nesting relies on a Box-Cox transformation with shape parameter λ to the conditional duration process and a possibly asymmetric shocks impact curve. We establish conditions for the existence of higher-order moments, strict stationarity, geometric ergodicity and β-mixing property with exponential decay. We next derive moment recursion relations and the autocovariance function of the power λ of the duration process. Finally, we assess the practical usefulness of our family of ACD models using NYSE transactions data, with special attention to IBM price durations. The results warrant the extra flexibility provided either by the Box-Cox transformation or by the asymmetric response to shocks.

A family of autoregressive conditional duration models

This paper develops a family of autoregressive conditional duration (ACD) models that encompasses most specifications in the literature. The nesting relies on a Box-Cox transformation with shape parameter λ to the conditional duration process and a possibly asymmetric shocks impact curve. We establish conditions for the existence of higher-order moments, strict stationarity, geometric ergodicity and β-mixing property with exponential decay. We next derive moment recursion relations and the autocovariance function of the power λ of the duration process. Finally, we assess the practical usefulness of our family of ACD models using NYSE price duration data on the IBM stock. The results warrant the extra flexibility provided either by the Box-Cox transformation or by the asymmetric response to shocks.

Non-linear gain - look-up table based approach for modeling a family of DC to DC converters based on transient response analysis

The purpose of this work is to propose a structure for simulating power systems using behavioral models of nonlinear DC to DC converters implemented through a look-up table of gains. This structure is specially designed for converters whose output impedance depends on the load current level, e.g. quasi-resonant converters. The proposed model is a generic one whose parameters can be obtained by direct measuring the transient response at different operating points. It also includes optional functionalities for modeling converters with current limitation and current sharing in paralleling characteristics. The pusposed structured also allows including aditional characteristics of the DC to DC converter as the efficency as a function of the input voltage and the output current or overvoltage and undervoltage protections. In addition, this proposed model is valid for overdamped and underdamped situations.

Modeling non-stationarity in intensity, duration and frequency of extreme rainfall over India

Significant changes are reported in extreme rainfall characteristics over India in recent studies though there are disagreements on the spatial uniformity and causes of trends. Based on recent theoretical advancements in the Extreme Value Theory (EVT), we analyze changes in extreme rainfall characteristics over India using a high-resolution daily gridded (1 degrees latitude x 1 degrees longitude) dataset. Intensity, duration and frequency of excess rain over a high threshold in the summer monsoon season are modeled by non-stationary distributions whose parameters vary with physical covariates like the El-Nino Southern Oscillation index (ENSO-index) which is an indicator of large-scale natural variability, global average temperature which is an indicator of human-induced global warming and local mean temperatures which possibly indicate more localized changes. Each non-stationary model considers one physical covariate and the best chosen statistical model at each rainfall grid gives the most significant physical driver for each extreme rainfall characteristic at that grid. Intensity, duration and frequency of extreme rainfall exhibit non-stationarity due to different drivers and no spatially uniform pattern is observed in the changes in them across the country. At most of the locations, duration of extreme rainfall spells is found to be stationary, while non-stationary associations between intensity and frequency and local changes in temperature are detected at a large number of locations. This study presents the first application of nonstationary statistical modeling of intensity, duration and frequency of extreme rainfall over India. The developed models are further used for rainfall frequency analysis to show changes in the 100-year extreme rainfall event. Our findings indicate the varying nature of each extreme rainfall characteristic and their drivers and emphasize the necessity of a comprehensive framework to assess resulting risks of precipitation induced flooding. (C) 2014 Elsevier B.V. All rights reserved.

Adjustment in children with intractable epilepsy:importance of seizure duration and family factors

Seventy-five families of children with intractable epilepsy but without a severe learning disability (mean age 7 years 1 month, SD 2 years 6 months; range 2 to 12 years) who attended a regional paediatric neurology service, were surveyed. A postal questionnaire was used which included standardized measures of child and family adjustment; forty-eight families responded (64%; 31 males, 17 females). There was no significant difference between responders and non-responders in terms of age, sex, number of other chronic illnesses and disabilities, age at epilepsy diagnosis, seizure type, nor number of antiepileptic drugs currently prescribed (p > 0.05). The importance of including multidimensional measures of outcome was highlighted by the finding that epilepsy, pharmacological, and psychosocial factors were differentially associated with specific adjustment difficulties. Two factors appeared to be most pervasively implicated across a range of adjustment problems: frequency of rectal diazepam administration and family patterns of relating to each other (p <0.05). It appeared that duration of seizures (as indicated by frequency of rectal diazepam administration), rather than the frequency of seizures per se, was more pernicious in terms of poor adjustment. Intrafamilial relations (degree of conflict/cohesion and soon) were not only associated with adjustment difficulties in the child, but also with the frequency of seizures themselves. Implications for psychological interventions in intractable epilepsy in childhood are highlighted.

Design, implementation and application of a generic framework for integrated regional land-use modeling

Land use is a crucial link between human activities and the natural environment and one of the main driving forces of global environmental change. Large parts of the terrestrial land surface are used for agriculture, forestry, settlements and infrastructure. Given the importance of land use, it is essential to understand the multitude of influential factors and resulting land use patterns. An essential methodology to study and quantify such interactions is provided by the adoption of land-use models. By the application of land-use models, it is possible to analyze the complex structure of linkages and feedbacks and to also determine the relevance of driving forces. Modeling land use and land use changes has a long-term tradition. In particular on the regional scale, a variety of models for different regions and research questions has been created. Modeling capabilities grow with steady advances in computer technology, which on the one hand are driven by increasing computing power on the other hand by new methods in software development, e.g. object- and component-oriented architectures. In this thesis, SITE (Simulation of Terrestrial Environments), a novel framework for integrated regional sland-use modeling, will be introduced and discussed. Particular features of SITE are the notably extended capability to integrate models and the strict separation of application and implementation. These features enable efficient development, test and usage of integrated land-use models. On its system side, SITE provides generic data structures (grid, grid cells, attributes etc.) and takes over the responsibility for their administration. By means of a scripting language (Python) that has been extended by language features specific for land-use modeling, these data structures can be utilized and manipulated by modeling applications. The scripting language interpreter is embedded in SITE. The integration of sub models can be achieved via the scripting language or by usage of a generic interface provided by SITE. Furthermore, functionalities important for land-use modeling like model calibration, model tests and analysis support of simulation results have been integrated into the generic framework. During the implementation of SITE, specific emphasis was laid on expandability, maintainability and usability. Along with the modeling framework a land use model for the analysis of the stability of tropical rainforest margins was developed in the context of the collaborative research project STORMA (SFB 552). In a research area in Central Sulawesi, Indonesia, socio-environmental impacts of land-use changes were examined. SITE was used to simulate land-use dynamics in the historical period of 1981 to 2002. Analogous to that, a scenario that did not consider migration in the population dynamics, was analyzed. For the calculation of crop yields and trace gas emissions, the DAYCENT agro-ecosystem model was integrated. In this case study, it could be shown that land-use changes in the Indonesian research area could mainly be characterized by the expansion of agricultural areas at the expense of natural forest. For this reason, the situation had to be interpreted as unsustainable even though increased agricultural use implied economic improvements and higher farmers' incomes. Due to the importance of model calibration, it was explicitly addressed in the SITE architecture through the introduction of a specific component. The calibration functionality can be used by all SITE applications and enables largely automated model calibration. Calibration in SITE is understood as a process that finds an optimal or at least adequate solution for a set of arbitrarily selectable model parameters with respect to an objective function. In SITE, an objective function typically is a map comparison algorithm capable of comparing a simulation result to a reference map. Several map optimization and map comparison methodologies are available and can be combined. The STORMA land-use model was calibrated using a genetic algorithm for optimization and the figure of merit map comparison measure as objective function. The time period for the calibration ranged from 1981 to 2002. For this period, respective reference land-use maps were compiled. It could be shown, that an efficient automated model calibration with SITE is possible. Nevertheless, the selection of the calibration parameters required detailed knowledge about the underlying land-use model and cannot be automated. In another case study decreases in crop yields and resulting losses in income from coffee cultivation were analyzed and quantified under the assumption of four different deforestation scenarios. For this task, an empirical model, describing the dependence of bee pollination and resulting coffee fruit set from the distance to the closest natural forest, was integrated. Land-use simulations showed, that depending on the magnitude and location of ongoing forest conversion, pollination services are expected to decline continuously. This results in a reduction of coffee yields of up to 18% and a loss of net revenues per hectare of up to 14%. However, the study also showed that ecological and economic values can be preserved if patches of natural vegetation are conservated in the agricultural landscape. -----------------------------------------------------------------------

Finite element modeling of a generic stemless hip implant design in comparison with conventional hip implants

In this study, the finite element modeling and comparison of the stress and strain analyses were carried out for three different structures that are intact bone, stemless implant and stemmed one. Currently proposed stemless design studied here is the generic concept of stemless implant. This generic stemless implant reconstruction was numerically compared to the conventional stemmed implant and also to the intact bone as control solution. Two loading conditions were applied to the most proximal part of the models, while the most distal part was fixed for all degrees of freedom. The models were divided into two regions and studied along two paths of medial and lateral aspect. The results of this study showed that the stemless implant had less deviation from the control solution of the bone in all regions and in both loading conditions, comparing to the large deviation of the stemmed implant from the intact bone. However, it was shown that the fixation of this type of implant and its effect on sub-trochanter region must be carefully considered for designing the final product of any specific design of stemless implant.