Data-driven Automatic Edge Sharpening

Sharpening is a powerful image transformation because sharp edges can bring out image details. Sharpness is achieved by increasing local contrast and reducing edge widths. We present a method that enhances sharpness of images and thereby their perceptual quality. Most existing enhancement techniques require user input to improve the perception of the scene in a manner most pleasing to the particular user. Our goal of image enhancement is to improve the perception of sharpness in digital images for human viewers. We consider two parameters in order to exaggerate the differences between local intensities. The two parameters exploit local contrast and widths of edges. We start from the assumption that color, texture, or objects of focus such as faces affect the human perception of photographs. When human raters are presented with a collection of images with different sharpness and asked to rank them according to perceived sharpness, the results have shown that there is a statistical consensus among the raters. We introduce a ramp enhancement technique by modifying the optimal overshoot in the ramp for different region contrasts as well as the new ramp width. Optimal parameter values are searched to be applied to regions under the criteria mentioned above. In this way, we aim to enhance digital images automatically to create pleasing image output for common users.

DATA DRIVEN APPROACHES TO IDENTIFY DETERMINANTS OF HEART DISEASES AND CANCER RESISTANCE

Cancer and cardio-vascular diseases are the leading causes of death world-wide. Caused by systemic genetic and molecular disruptions in cells, these disorders are the manifestation of profound disturbance of normal cellular homeostasis. People suffering or at high risk for these disorders need early diagnosis and personalized therapeutic intervention. Successful implementation of such clinical measures can significantly improve global health. However, development of effective therapies is hindered by the challenges in identifying genetic and molecular determinants of the onset of diseases; and in cases where therapies already exist, the main challenge is to identify molecular determinants that drive resistance to the therapies. Due to the progress in sequencing technologies, the access to a large genome-wide biological data is now extended far beyond few experimental labs to the global research community. The unprecedented availability of the data has revolutionized the capabilities of computational researchers, enabling them to collaboratively address the long standing problems from many different perspectives. Likewise, this thesis tackles the two main public health related challenges using data driven approaches. Numerous association studies have been proposed to identify genomic variants that determine disease. However, their clinical utility remains limited due to their inability to distinguish causal variants from associated variants. In the presented thesis, we first propose a simple scheme that improves association studies in supervised fashion and has shown its applicability in identifying genomic regulatory variants associated with hypertension. Next, we propose a coupled Bayesian regression approach -- eQTeL, which leverages epigenetic data to estimate regulatory and gene interaction potential, and identifies combinations of regulatory genomic variants that explain the gene expression variance. On human heart data, eQTeL not only explains a significantly greater proportion of expression variance in samples, but also predicts gene expression more accurately than other methods. We demonstrate that eQTeL accurately detects causal regulatory SNPs by simulation, particularly those with small effect sizes. Using various functional data, we show that SNPs detected by eQTeL are enriched for allele-specific protein binding and histone modifications, which potentially disrupt binding of core cardiac transcription factors and are spatially proximal to their target. eQTeL SNPs capture a substantial proportion of genetic determinants of expression variance and we estimate that 58% of these SNPs are putatively causal. The challenge of identifying molecular determinants of cancer resistance so far could only be dealt with labor intensive and costly experimental studies, and in case of experimental drugs such studies are infeasible. Here we take a fundamentally different data driven approach to understand the evolving landscape of emerging resistance. We introduce a novel class of genetic interactions termed synthetic rescues (SR) in cancer, which denotes a functional interaction between two genes where a change in the activity of one vulnerable gene (which may be a target of a cancer drug) is lethal, but subsequently altered activity of its partner rescuer gene restores cell viability. Next we describe a comprehensive computational framework --termed INCISOR-- for identifying SR underlying cancer resistance. Applying INCISOR to mine The Cancer Genome Atlas (TCGA), a large collection of cancer patient data, we identified the first pan-cancer SR networks, composed of interactions common to many cancer types. We experimentally test and validate a subset of these interactions involving the master regulator gene mTOR. We find that rescuer genes become increasingly activated as breast cancer progresses, testifying to pervasive ongoing rescue processes. We show that SRs can be utilized to successfully predict patients' survival and response to the majority of current cancer drugs, and importantly, for predicting the emergence of drug resistance from the initial tumor biopsy. Our analysis suggests a potential new strategy for enhancing the effectiveness of existing cancer therapies by targeting their rescuer genes to counteract resistance. The thesis provides statistical frameworks that can harness ever increasing high throughput genomic data to address challenges in determining the molecular underpinnings of hypertension, cardiovascular disease and cancer resistance. We discover novel molecular mechanistic insights that will advance the progress in early disease prevention and personalized therapeutics. Our analyses sheds light on the fundamental biological understanding of gene regulation and interaction, and opens up exciting avenues of translational applications in risk prediction and therapeutics.

Parcellation of fMRI datasets with ICA and PLS: a data driven approach

Inter-subject parcellation of functional Magnetic Resonance Imaging (fMRI) data based on a standard General Linear Model (GLM) and spectral clustering was recently proposed as a means to alleviate the issues associated with spatial normalization in fMRI. However, for all its appeal, a GLM-based parcellation approach introduces its own biases, in the form of a priori knowledge about the shape of Hemodynamic Response Function (HRF) and task-related signal changes, or about the subject behaviour during the task. In this paper, we introduce a data-driven version of the spectral clustering parcellation, based on Independent Component Analysis (ICA) and Partial Least Squares (PLS) instead of the GLM. First, a number of independent components are automatically selected. Seed voxels are then obtained from the associated ICA maps and we compute the PLS latent variables between the fMRI signal of the seed voxels (which covers regional variations of the HRF) and the principal components of the signal across all voxels. Finally, we parcellate all subjects data with a spectral clustering of the PLS latent variables. We present results of the application of the proposed method on both single-subject and multi-subject fMRI datasets. Preliminary experimental results, evaluated with intra-parcel variance of GLM t-values and PLS derived t-values, indicate that this data-driven approach offers improvement in terms of parcellation accuracy over GLM based techniques.

DATA DRIVEN INTELLIGENT AGENT NETWORKS FOR ADAPTIVE MONITORING AND CONTROL

To analyze the characteristics and predict the dynamic behaviors of complex systems over time, comprehensive research to enable the development of systems that can intelligently adapt to the evolving conditions and infer new knowledge with algorithms that are not predesigned is crucially needed. This dissertation research studies the integration of the techniques and methodologies resulted from the fields of pattern recognition, intelligent agents, artificial immune systems, and distributed computing platforms, to create technologies that can more accurately describe and control the dynamics of real-world complex systems. The need for such technologies is emerging in manufacturing, transportation, hazard mitigation, weather and climate prediction, homeland security, and emergency response. Motivated by the ability of mobile agents to dynamically incorporate additional computational and control algorithms into executing applications, mobile agent technology is employed in this research for the adaptive sensing and monitoring in a wireless sensor network. Mobile agents are software components that can travel from one computing platform to another in a network and carry programs and data states that are needed for performing the assigned tasks. To support the generation, migration, communication, and management of mobile monitoring agents, an embeddable mobile agent system (Mobile-C) is integrated with sensor nodes. Mobile monitoring agents visit distributed sensor nodes, read real-time sensor data, and perform anomaly detection using the equipped pattern recognition algorithms. The optimal control of agents is achieved by mimicking the adaptive immune response and the application of multi-objective optimization algorithms. The mobile agent approach provides potential to reduce the communication load and energy consumption in monitoring networks. The major research work of this dissertation project includes: (1) studying effective feature extraction methods for time series measurement data; (2) investigating the impact of the feature extraction methods and dissimilarity measures on the performance of pattern recognition; (3) researching the effects of environmental factors on the performance of pattern recognition; (4) integrating an embeddable mobile agent system with wireless sensor nodes; (5) optimizing agent generation and distribution using artificial immune system concept and multi-objective algorithms; (6) applying mobile agent technology and pattern recognition algorithms for adaptive structural health monitoring and driving cycle pattern recognition; (7) developing a web-based monitoring network to enable the visualization and analysis of real-time sensor data remotely. Techniques and algorithms developed in this dissertation project will contribute to research advances in networked distributed systems operating under changing environments.

Learning data-driven models of non-verbal behaviors for building rapport using an intelligent virtual agent

There is a growing societal need to address the increasing prevalence of behavioral health issues, such as obesity, alcohol or drug use, and general lack of treatment adherence for a variety of health problems. The statistics, worldwide and in the USA, are daunting. Excessive alcohol use is the third leading preventable cause of death in the United States (with 79,000 deaths annually), and is responsible for a wide range of health and social problems. On the positive side though, these behavioral health issues (and associated possible diseases) can often be prevented with relatively simple lifestyle changes, such as losing weight with a diet and/or physical exercise, or learning how to reduce alcohol consumption. Medicine has therefore started to move toward finding ways of preventively promoting wellness, rather than solely treating already established illness.^ Evidence-based patient-centered Brief Motivational Interviewing (BMI) interventions have been found particularly effective in helping people find intrinsic motivation to change problem behaviors after short counseling sessions, and to maintain healthy lifestyles over the long-term. Lack of locally available personnel well-trained in BMI, however, often limits access to successful interventions for people in need. To fill this accessibility gap, Computer-Based Interventions (CBIs) have started to emerge. Success of the CBIs, however, critically relies on insuring engagement and retention of CBI users so that they remain motivated to use these systems and come back to use them over the long term as necessary.^ Because of their text-only interfaces, current CBIs can therefore only express limited empathy and rapport, which are the most important factors of health interventions. Fortunately, in the last decade, computer science research has progressed in the design of simulated human characters with anthropomorphic communicative abilities. Virtual characters interact using humans’ innate communication modalities, such as facial expressions, body language, speech, and natural language understanding. By advancing research in Artificial Intelligence (AI), we can improve the ability of artificial agents to help us solve CBI problems.^ To facilitate successful communication and social interaction between artificial agents and human partners, it is essential that aspects of human social behavior, especially empathy and rapport, be considered when designing human-computer interfaces. Hence, the goal of the present dissertation is to provide a computational model of rapport to enhance an artificial agent’s social behavior, and to provide an experimental tool for the psychological theories shaping the model. Parts of this thesis were already published in [LYL+12, AYL12, AL13, ALYR13, LAYR13, YALR13, ALY14].^

A MapReduce-based nearest neighbor approach for big-data-driven traffic flow prediction

In big-data-driven traffic flow prediction systems, the robustness of prediction performance depends on accuracy and timeliness. This paper presents a new MapReduce-based nearest neighbor (NN) approach for traffic flow prediction using correlation analysis (TFPC) on a Hadoop platform. In particular, we develop a real-time prediction system including two key modules, i.e., offline distributed training (ODT) and online parallel prediction (OPP). Moreover, we build a parallel k-nearest neighbor optimization classifier, which incorporates correlation information among traffic flows into the classification process. Finally, we propose a novel prediction calculation method, combining the current data observed in OPP and the classification results obtained from large-scale historical data in ODT, to generate traffic flow prediction in real time. The empirical study on real-world traffic flow big data using the leave-one-out cross validation method shows that TFPC significantly outperforms four state-of-the-art prediction approaches, i.e., autoregressive integrated moving average, Naïve Bayes, multilayer perceptron neural networks, and NN regression, in terms of accuracy, which can be improved 90.07% in the best case, with an average mean absolute percent error of 5.53%. In addition, it displays excellent speedup, scaleup, and sizeup.

Comparison of ANN models for predicting water quality in distribution systems

Deterministic models have been widely used to predict water quality in distribution systems, but their calibration requires extensive and accurate data sets for numerous parameters. In this study, alternative data-driven modeling approaches based on artificial neural networks (ANNs) were used to predict temporal variations of two important characteristics of water quality chlorine residual and biomass concentrations. The authors considered three types of ANN algorithms. Of these, the Levenberg-Marquardt algorithm provided the best results in predicting residual chlorine and biomass with error-free and ``noisy'' data. The ANN models developed here can generate water quality scenarios of piped systems in real time to help utilities determine weak points of low chlorine residual and high biomass concentration and select optimum remedial strategies.

Adaptive Algorithms for Personalized Diabetes Treatment

Dynamic systems, especially in real-life applications, are often determined by inter-/intra-variability, uncertainties and time-varying components. Physiological systems are probably the most representative example in which population variability, vital signal measurement noise and uncertain dynamics render their explicit representation and optimization a rather difficult task. Systems characterized by such challenges often require the use of adaptive algorithmic solutions able to perform an iterative structural and/or parametrical update process towards optimized behavior. Adaptive optimization presents the advantages of (i) individualization through learning of basic system characteristics, (ii) ability to follow time-varying dynamics and (iii) low computational cost. In this chapter, the use of online adaptive algorithms is investigated in two basic research areas related to diabetes management: (i) real-time glucose regulation and (ii) real-time prediction of hypo-/hyperglycemia. The applicability of these methods is illustrated through the design and development of an adaptive glucose control algorithm based on reinforcement learning and optimal control and an adaptive, personalized early-warning system for the recognition and alarm generation against hypo- and hyperglycemic events.

Linear latent force models using Gaussian Processes

Purely data-driven approaches for machine learning present difficulties when data are scarce relative to the complexity of the model or when the model is forced to extrapolate. On the other hand, purely mechanistic approaches need to identify and specify all the interactions in the problem at hand (which may not be feasible) and still leave the issue of how to parameterize the system. In this paper, we present a hybrid approach using Gaussian processes and differential equations to combine data-driven modeling with a physical model of the system. We show how different, physically inspired, kernel functions can be developed through sensible, simple, mechanistic assumptions about the underlying system. The versatility of our approach is illustrated with three case studies from motion capture, computational biology, and geostatistics.

Degradation modeling and monitoring of machines using operation-specific hidden Markov models

In this paper, a novel data-driven approach to monitoring of systems operating under variable operating conditions is described. The method is based on characterizing the degradation process via a set of operation-specific hidden Markov models (HMMs), whose hidden states represent the unobservable degradation states of the monitored system while its observable symbols represent the sensor readings. Using the HMM framework, modeling, identification and monitoring methods are detailed that allow one to identify a HMM of degradation for each operation from mixed-operation data and perform operation-specific monitoring of the system. Using a large data set provided by a major manufacturer, the new methods are applied to a semiconductor manufacturing process running multiple operations in a production environment.

Generalized Laplacian Eigenmaps for Modeling and Tracking Human Motions

This paper presents generalized Laplacian eigenmaps, a novel dimensionality reduction approach designed to address stylistic variations in time series. It generates compact and coherent continuous spaces whose geometry is data-driven. This paper also introduces graph-based particle filter, a novel methodology conceived for efficient tracking in low dimensional space derived from a spectral dimensionality reduction method. Its strengths are a propagation scheme, which facilitates the prediction in time and style, and a noise model coherent with the manifold, which prevents divergence, and increases robustness. Experiments show that a combination of both techniques achieves state-of-the-art performance for human pose tracking in underconstrained scenarios.

Modeling and optimization of extracellular polysaccharides production by Enterobacter A47

Polysaccharides are gaining increasing attention as potential environmental friendly and sustainable building blocks in many fields of the (bio)chemical industry. The microbial production of polysaccharides is envisioned as a promising path, since higher biomass growth rates are possible and therefore higher productivities may be achieved compared to vegetable or animal polysaccharides sources. This Ph.D. thesis focuses on the modeling and optimization of a particular microbial polysaccharide, namely the production of extracellular polysaccharides (EPS) by the bacterial strain Enterobacter A47. Enterobacter A47 was found to be a metabolically versatile organism in terms of its adaptability to complex media, notably capable of achieving high growth rates in media containing glycerol byproduct from the biodiesel industry. However, the industrial implementation of this production process is still hampered due to a largely unoptimized process. Kinetic rates from the bioreactor operation are heavily dependent on operational parameters such as temperature, pH, stirring and aeration rate. The increase of culture broth viscosity is a common feature of this culture and has a major impact on the overall performance. This fact complicates the mathematical modeling of the process, limiting the possibility to understand, control and optimize productivity. In order to tackle this difficulty, data-driven mathematical methodologies such as Artificial Neural Networks can be employed to incorporate additional process data to complement the known mathematical description of the fermentation kinetics. In this Ph.D. thesis, we have adopted such an hybrid modeling framework that enabled the incorporation of temperature, pH and viscosity effects on the fermentation kinetics in order to improve the dynamical modeling and optimization of the process. A model-based optimization method was implemented that enabled to design bioreactor optimal control strategies in the sense of EPS productivity maximization. It is also critical to understand EPS synthesis at the level of the bacterial metabolism, since the production of EPS is a tightly regulated process. Methods of pathway analysis provide a means to unravel the fundamental pathways and their controls in bioprocesses. In the present Ph.D. thesis, a novel methodology called Principal Elementary Mode Analysis (PEMA) was developed and implemented that enabled to identify which cellular fluxes are activated under different conditions of temperature and pH. It is shown that differences in these two parameters affect the chemical composition of EPS, hence they are critical for the regulation of the product synthesis. In future studies, the knowledge provided by PEMA could foster the development of metabolically meaningful control strategies that target the EPS sugar content and oder product quality parameters.

Parameterization Of Turbulence Models Using 3DVAR Data Assimilation

In this research the 3DVAR data assimilation scheme is implemented in the numerical model DIVAST in order to optimize the performance of the numerical model by selecting an appropriate turbulence scheme and tuning its parameters. Two turbulence closure schemes: the Prandtl mixing length model and the two-equation k-ε model were incorporated into DIVAST and examined with respect to their universality of application, complexity of solutions, computational efficiency and numerical stability. A square harbour with one symmetrical entrance subject to tide-induced flows was selected to investigate the structure of turbulent flows. The experimental part of the research was conducted in a tidal basin. A significant advantage of such laboratory experiment is a fully controlled environment where domain setup and forcing are user-defined. The research shows that the Prandtl mixing length model and the two-equation k-ε model, with default parameterization predefined according to literature recommendations, overestimate eddy viscosity which in turn results in a significant underestimation of velocity magnitudes in the harbour. The data assimilation of the model-predicted velocity and laboratory observations significantly improves model predictions for both turbulence models by adjusting modelled flows in the harbour to match de-errored observations. 3DVAR allows also to identify and quantify shortcomings of the numerical model. Such comprehensive analysis gives an optimal solution based on which numerical model parameters can be estimated. The process of turbulence model optimization by reparameterization and tuning towards optimal state led to new constants that may be potentially applied to complex turbulent flows, such as rapidly developing flows or recirculating flows.