Dynamic Programming Approaches for Estimating and Applying Large-scale Discrete Choice Models

People go through their life making all kinds of decisions, and some of these decisions affect their demand for transportation, for example, their choices of where to live and where to work, how and when to travel and which route to take. Transport related choices are typically time dependent and characterized by large number of alternatives that can be spatially correlated. This thesis deals with models that can be used to analyze and predict discrete choices in large-scale networks. The proposed models and methods are highly relevant for, but not limited to, transport applications. We model decisions as sequences of choices within the dynamic discrete choice framework, also known as parametric Markov decision processes. Such models are known to be difficult to estimate and to apply to make predictions because dynamic programming problems need to be solved in order to compute choice probabilities. In this thesis we show that it is possible to explore the network structure and the flexibility of dynamic programming so that the dynamic discrete choice modeling approach is not only useful to model time dependent choices, but also makes it easier to model large-scale static choices. The thesis consists of seven articles containing a number of models and methods for estimating, applying and testing large-scale discrete choice models. In the following we group the contributions under three themes: route choice modeling, large-scale multivariate extreme value (MEV) model estimation and nonlinear optimization algorithms. Five articles are related to route choice modeling. We propose different dynamic discrete choice models that allow paths to be correlated based on the MEV and mixed logit models. The resulting route choice models become expensive to estimate and we deal with this challenge by proposing innovative methods that allow to reduce the estimation cost. For example, we propose a decomposition method that not only opens up for possibility of mixing, but also speeds up the estimation for simple logit models, which has implications also for traffic simulation. Moreover, we compare the utility maximization and regret minimization decision rules, and we propose a misspecification test for logit-based route choice models. The second theme is related to the estimation of static discrete choice models with large choice sets. We establish that a class of MEV models can be reformulated as dynamic discrete choice models on the networks of correlation structures. These dynamic models can then be estimated quickly using dynamic programming techniques and an efficient nonlinear optimization algorithm. Finally, the third theme focuses on structured quasi-Newton techniques for estimating discrete choice models by maximum likelihood. We examine and adapt switching methods that can be easily integrated into usual optimization algorithms (line search and trust region) to accelerate the estimation process. The proposed dynamic discrete choice models and estimation methods can be used in various discrete choice applications. In the area of big data analytics, models that can deal with large choice sets and sequential choices are important. Our research can therefore be of interest in various demand analysis applications (predictive analytics) or can be integrated with optimization models (prescriptive analytics). Furthermore, our studies indicate the potential of dynamic programming techniques in this context, even for static models, which opens up a variety of future research directions.

Uma arquitetura de referência para o processamento distribuído de stream de dados em soluções analíticas de near real-time

Dissertação (mestrado)—Universidade de Brasília, Faculdade de Tecnologia, Departamento de Engenharia Elétrica, 2015.

Identity, addressing, authenticity and audit requirements for trust in ERP, analytics and big/open data in a “Cloud” computing environment : a review and proposal

Enterprises, both public and private, have rapidly commenced using the benefits of enterprise resource planning (ERP) combined with business analytics and “open data sets” which are often outside the control of the enterprise to gain further efficiencies, build new service operations and increase business activity. In many cases, these business activities are based around relevant software systems hosted in a “cloud computing” environment. “Garbage in, garbage out”, or “GIGO”, is a term long used to describe problems in unqualified dependency on information systems, dating from the 1960s. However, a more pertinent variation arose sometime later, namely “garbage in, gospel out” signifying that with large scale information systems, such as ERP and usage of open datasets in a cloud environment, the ability to verify the authenticity of those data sets used may be almost impossible, resulting in dependence upon questionable results. Illicit data set “impersonation” becomes a reality. At the same time the ability to audit such results may be an important requirement, particularly in the public sector. This paper discusses the need for enhancement of identity, reliability, authenticity and audit services, including naming and addressing services, in this emerging environment and analyses some current technologies that are offered and which may be appropriate. However, severe limitations to addressing these requirements have been identified and the paper proposes further research work in the area.

Clinical data warehousing : a business analytics approach for managing health data

Heterogeneous health data is a critical issue when managing health information for quality decision making processes. In this paper we examine the efficient aggregation of lifestyle information through a data warehousing architecture lens. We present a proof of concept for a clinical data warehouse architecture that enables evidence based decision making processes by integrating and organising disparate data silos in support of healthcare services improvement paradigms.

Visual analytics for large-scale bioinformatic data sets

Rapid advances in sequencing technologies (Next Generation Sequencing or NGS) have led to a vast increase in the quantity of bioinformatics data available, with this increasing scale presenting enormous challenges to researchers seeking to identify complex interactions. This paper is concerned with the domain of transcriptional regulation, and the use of visualisation to identify relationships between specific regulatory proteins (the transcription factors or TFs) and their associated target genes (TGs). We present preliminary work from an ongoing study which aims to determine the effectiveness of different visual representations and large scale displays in supporting discovery. Following an iterative process of implementation and evaluation, representations were tested by potential users in the bioinformatics domain to determine their efficacy, and to understand better the range of ad hoc practices among bioinformatics literate users. Results from two rounds of small scale user studies are considered with initial findings suggesting that bioinformaticians require richly detailed views of TF data, features to compare TF layouts between organisms quickly, and ways to keep track of interesting data points.

Making sense of game based user data: learning analytics in applied games

Digital learning games are useful educational tools with high motivational potential. With the application of games for instruction there comes the need of acknowledging learning game experiences also in the context of educational assessment. Learning analytics provides new opportunities for supporting assessment in and of educational games. We give an overview of current learning analytics methods in this field and reflect on existing challenges. An approach of providing reusable software assets for interaction assessment and evaluation in games is presented. This is part of a broader initiative of making available advanced methodologies and tools for supporting applied game development.

Topographic visual analytics of multibeam dynamic SONAR data

This paper considers the problem of low-dimensional visualisation of very high dimensional information sources for the purpose of situation awareness in the maritime environment. In response to the requirement for human decision support aids to reduce information overload (and specifically, data amenable to inter-point relative similarity measures) appropriate to the below-water maritime domain, we are investigating a preliminary prototype topographic visualisation model. The focus of the current paper is on the mathematical problem of exploiting a relative dissimilarity representation of signals in a visual informatics mapping model, driven by real-world sonar systems. A realistic noise model is explored and incorporated into non-linear and topographic visualisation algorithms building on the approach of [9]. Concepts are illustrated using a real world dataset of 32 hydrophones monitoring a shallow-water environment in which targets are present and dynamic.

Using knowledge management to give context to analytics and big data and reduce strategic risk

At the moment, the phrases “big data” and “analytics” are often being used as if they were magic incantations that will solve all an organization’s problems at a stroke. The reality is that data on its own, even with the application of analytics, will not solve any problems. The resources that analytics and big data can consume represent a significant strategic risk if applied ineffectively. Any analysis of data needs to be guided, and to lead to action. So while analytics may lead to knowledge and intelligence (in the military sense of that term), it also needs the input of knowledge and intelligence (in the human sense of that term). And somebody then has to do something new or different as a result of the new insights, or it won’t have been done to any purpose. Using an analytics example concerning accounts payable in the public sector in Canada, this paper reviews thinking from the domains of analytics, risk management and knowledge management, to show some of the pitfalls, and to present a holistic picture of how knowledge management might help tackle the challenges of big data and analytics.

Visual analytics of medical and biological data

Otto-von-Guericke-Universität Magdeburg, Fakultät für Informatik, Habilitationsschrift, 2016

Data mining for lifetime prediction of metallic components

The ability to accurately predict the lifetime of building components is crucial to optimizing building design, material selection and scheduling of required maintenance. This paper discusses a number of possible data mining methods that can be applied to do the lifetime prediction of metallic components and how different sources of service life information could be integrated to form the basis of the lifetime prediction model

An innovative information gathering and data analysis platform for railway level crossing safety data

This paper describes an innovative platform that facilitates the collection of objective safety data around occurrences at railway level crossings using data sources including forward-facing video, telemetry from trains and geo-referenced asset and survey data. This platform is being developed with support by the Australian rail industry and the Cooperative Research Centre for Rail Innovation. The paper provides a description of the underlying accident causation model, the development methodology and refinement process as well as a description of the data collection platform. The paper concludes with a brief discussion of benefits this project is expected to provide the Australian rail industry.

The multi-platform user engagement and data visualisation tool

The ability to identify and assess user engagement with transmedia productions is vital to the success of individual projects and the sustainability of this mode of media production as a whole. It is essential that industry players have access to tools and methodologies that offer the most complete and accurate picture of how audiences/users engage with their productions and which assets generate the most valuable returns of investment. Drawing upon research conducted with Hoodlum Entertainment, a Brisbane-based transmedia producer, this project involved an initial assessment of the way engagement tends to be understood, why standard web analytics tools are ill-suited to measuring it, how a customised tool could offer solutions, and why this question of measuring engagement is so vital to the future of transmedia as a sustainable industry. Working with data provided by Hoodlum Entertainment and Foxtel Marketing, the outcome of the study was a prototype for a custom data visualisation tool that allowed access, manipulation and presentation of user engagement data, both historic and predictive. The prototyped interfaces demonstrate how the visualization tool would collect and organise data specific to multiplatform projects by aggregating data across a number of platform reporting tools. Such a tool is designed to encompass not only platforms developed by the transmedia producer but also sites developed by fans. This visualisation tool accounted for multiplatform experience projects whose top level is comprised of people, platforms and content. People include characters, actors, audience, distributors and creators. Platforms include television, Facebook and other relevant social networks, literature, cinema and other media that might be included in the multiplatform experience. Content refers to discreet media texts employed within the platform, such as tweet, a You Tube video, a Facebook post, an email, a television episode, etc. Core content is produced by the creators’ multiplatform experiences to advance the narrative, while complimentary content generated by audience members offers further contributions to the experience. Equally important is the timing with which the components of the experience are introduced and how they interact with and impact upon each other. Being able to combine, filter and sort these elements in multiple ways we can better understand the value of certain components of a project. It also offers insights into the relationship between the timing of the release of components and user activity associated with them, which further highlights the efficacy (or, indeed, failure) of assets as catalysts for engagement. In collaboration with Hoodlum we have developed a number of design scenarios experimenting with the ways in which data can be visualised and manipulated to tell a more refined story about the value of user engagement with certain project components and activities. This experimentation will serve as the basis for future research.