Search NASA⌕ Search

SEARCH · Search NASA

Results for “Data Mining”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 541 records · Page 30

Distributed non-negative matrix factorization with determination of the number of latent features

The holistic analysis and understanding of the latent (that is, not directly observable) variables and patterns buried in large datasets is crucial for data-driven science, decision making and emergency response. Such exploratory analyses require devising unsupervised learning methods for data mining and extraction of the latent features, and non-negative matrix factorization (NMF) is one of the prominent such methods. NMF is based on compute-intense non-convex constrained minimization, which, for large datasets requires fast and distributed algorithms. However, current parallel implementations of NMF fail to estimate the number of latent features. In practice, identifying these features is both difficult and significant for pattern recognition and latent feature analysis, especially for large dense matrices. Here, we introduce a distributed NMF algorithm coupled with distributed custom clustering followed by a stability analysis on dense data, which we call DnMFk, to determine the number of latent variables. The results on synthetic data and the classical Swimmer data set demonstrate the accuracy of model determination while scaling nearly linearly across multiple processors for large data. Further, we employ DnMFk to determine the number of hidden features from a terabyte matrix.

97 MATHEMATICS AND COMPUTING↗

Physics of Boundaries and their Interactions in Space Plasmas

This final report describes a brief summary of our accomplishments during the complete contract period. Traditionally, due to computational limitations, it has been impossible to obtain a global view of the magnetosphere on ion time and spatial scales. As a result, kinetic simulations have concentrated on the local structure of different magnetospheric discontinuities and boundaries. However, due to the emergence of low cost desktop superconductors, as well as by taking full advantage of latest advances in data mining and visualization technology, we were able to bypass our planned (proposed) regional simulations and proceed to large-scale 3-D and 2-D global hybrid simulations of the magnetosphere. As a result, although we are only finishing the second year of the proposed activity, much of the original scientific objectives have been surpassed and new avenues of investigation have been opened. Such simulations have led us to possible explanations of some long-standing issues in magnetospheric physics. They have also enabled us to make a number of important discoveries/predictions, which need to be looked for in satellite data. Examples include: (1) the finding that the bow shock can become unstable to the Kelvin-Helmholtz (KH;) (2) the discovery of a mechanism for intermittent reconnection due to ion physics which may be relevant to the explanation of the recurrence rate of flux transfer events (FTEs;) and (3) the finding that the current sheet in the near-Earth magnetotail region can become unstable to KH with detectable, unique ionospheric signatures. Further, we demonstrated a viable mechanism for the onset of reconnection at the magnetopause, examined the detailed structure of the boundary layer incorporating curvature effects, and provided an explanation for the large core fields observed within FTEs as well as flux ropes in the magnetotail.

Omidi, Nojan↗

Physics of Boundaries and their Interactions in Space Plasmas

This final report describes a brief summary of our accomplishments during the complete contract period. Traditionally, due to computational limitations, it has been impossible to obtain a global view of the magnetosphere on ion time and spatial scales. As a result, kinetic-simulations have concentrated on the local structure of different magnetospheric discontinuities and boundaries. However, due to the emergence of low cost supercomputers, as well as by taking full advantage of latest advances in data mining and visualization technology, we were able to bypass our planned (proposed) regional simulations and proceed to large-scale 3-D and 2-D global hybrid simulations of the magnetosphere. As a result, although we are only finishing the second year of the proposed activity, much of the original scientific objectives have been surpassed and new avenues of investigation have been opened. Such simulations have led us to possible explanations of some long-standing issues in magnetospheric physics. They have also enables us to make a number of important discoveries predictions, which need to be looked for in satellite data. Examples include the finding that the bow shock can become unstable to the Kelvin-Helmholtz (KH), (2) the discovery of a mechanism for intermittent reconnection due to ion physics which may be relevant to the explanation of the recurrence rate of flux transfer events (FTEs), and (3) this finding that the current sheet in the near-Earth magnetotail region can become unstable to KH with detectable, unique ionospheric signatures. Further, we demonstrated a viable mechanism for the onset of reconnection at the magnetopause, examined the detailed structure of the boundary layer incorporating curvature effects, and provided an explanation for the large core fields observed within FTEs as well as flux ropes in the magnetotail.

Omidi, Nojan↗

NFκB mediated elevation of KCNJ11 promotes tumor progression of hepatocellular carcinoma through interaction of lactate dehydrogenase A

Highlights: • Upregulated KCNJ11 predicts a poor prognosis and is regulated by NFκB signaling in HCC. • KCNJ11 knockdown suppresses HCC cell proliferation, apoptosis-resistance, and migration in vitro. • KCNJ11 interacts with LDHA. • LDHA partially mediated the oncogenic roles of KCNJ11 in HCC. It has been well documented that changes in ion fluxes across cellular membranes is fundamental in maintaining cellular homeostasis. Dysregulation and/or malfunction of ion channels are critical events in the pathogenesis of diverse diseases, including cancers. In this study, we focused on the study of K{sup +} channels in hepatocellular carcinoma (HCC). By data mining TCGA cohort, the expression of 27 K{sup +} channels was investigated and KCNJ11 was identified as a key dysregulated K{sup +} channels in HCC. KCNJ11 was differentially expressed in HCC and predicted a poor prognosis in HCC patients. Inhibition of NFκB signaling suppressed KCNJ11 expression in HCC cells. Knockdown of KCNJ11 expression inhibited cell proliferation, promoted cell apoptosis, and reduced cell invasive capacity. Mechanistically, we found that KCNJ11 promotes tumor progression through interaction with LDHA and enhancing its enzymatic activity. Pharmacological inhibition of LDHA largely compromised the oncogenic function of KCNJ11 in cell proliferation, cell apoptosis, and cell invasion. Collectively, our data, as a proof of principle, demonstrate that KCNJ11 acts as an oncogene in HCC though forming a complex with LDHA and suggest that targeting KCNJ11 can be developed as a candidate tool to dampen HCC.

60 APPLIED LIFE SCIENCES↗

CORE: A Global Aggregation Service for Open Access Papers

This paper introduces CORE, a widely used scholarly service, which provides access to the world’s largest collection of open access research publications, acquired from a global network of repositories and journals. CORE was created with the goal of enabling text and data mining of scientific literature and thus supporting scientific discovery, but it is now used in a wide range of use cases within higher education, industry, not-for-profit organisations, as well as by the general public. Through the provided services, CORE powers innovative use cases, such as plagiarism detection, in market-leading third-party organisations. CORE has played a pivotal role in the global move towards universal open access by making scientific knowledge more easily and freely discoverable. In this paper, we describe CORE’s continuously growing dataset and the motivation behind its creation, present the challenges associated with systematically gathering research papers from thousands of data providers worldwide at scale, and introduce the novel solutions that were developed to overcome these challenges. The paper then provides an in-depth discussion of the services and tools built on top of the aggregated data and finally examines several use cases that have leveraged the CORE dataset and services.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Machine Learning for Computational Heterogeneous Catalysis

Abstract Big data and artificial intelligence has revolutionized science in almost every field – from economics to physics. In the area of materials science and computational heterogeneous catalysis, this revolution has led to the development of scientific data repositories, as well as data mining and machine learning tools to investigate the vast materials space. The goal of using these tools is to establish a deeper understanding of the relations between materials properties and activity, selectivity and stability – the important figures of merit in catalysis. Based on these insights, catalyst design principles can be established, which hopefully lead us to discover highly efficient catalysts to solve pressing issues for a sustainable future and the synthesis of highly functional materials, chemicals and pharmaceuticals. The inherent complexity of catalytic reactions quests for machine learning methods to efficiently navigate through the high‐dimensional hyper‐surfaces in structure optimization problems to determine relevant chemical structures and transition states. In this review, we show how cutting edge data infrastructures and machine learning methods are being used to address problems in computational heterogeneous catalysis.

Schlexer Lamoureux, Philomena↗

Aquaporin 3 facilitates tumor growth in pancreatic cancer by modulating mTOR signaling

Aquaporins (AQP) have been demonstrated to be dysregulated in many human cancers and is thought to be involved in pancreatic carcinogenesis and progression. However, the oncogenic roles and underlying mechanism of AQP in pancreatic ductal adenocarcinoma (PDAC) remain largely unknown. In this study, by data mining of TCGA dataset and CCLE database, we identified that AQP3 is the major AQP expressed in PDAC. Then, the microRNA-874, was demonstrated to be a key regulator of AQP3 expression in PDAC cells. Genetic silencing of AQP3 expression had pronounced effects on cell proliferation and apoptosis of the PDAC cell lines BXPC3 and HPAFII. Introduction of microRNA-874 suppressed cell proliferation and promoted cell apoptosis, whereas inhibition of microRNA-874 had the opposite effect. Mechanistically, by a large-scale proteomic analysis, we revealed that AQP3 was significantly associated the activity of mTOR signaling. Moreover, modulation of AQP3 or microRNA-874 altered mTOR activity as demonstrated by the phosphorylation level of mTOR and its downstream target S6. Taken together, our data, as a proof of principle, suggest that AQP3 can promote tumor growth of pancreatic cancer cells by activating the Mtor signaling pathway and provide a potential therapeutic target in the treatment of PDAC. - Highlights: • Dysregulated expression of AQP3 is induced by miR-874 in pancreatic cancer. • miR-874-AQP3 axis is involved in cell proliferation and apoptosis of pancreatic cancer cells. • AQP3 correlates increased mTOR activity in pancreatic cancer.

60 APPLIED LIFE SCIENCES↗

Predicting Execution Times for Disk-based and In-Situ Parallel Data Analytics (Final Technical Report)

In recent years, there has been a significant amount of interests in in-situ analytics on simulation programs. For a variety of reasons, it is desirable to be able to predict the execution time of an analytics program. At the same time, frameworks such as MapReduce have become popular for scientific data analytics. This paper focuses on developing performance models for predicting execution time of parallel data analytics, with a special emphasis on in-situ analytics. We take two distinct approach towards performance prediction. We first expand SKOPE (a SKeleton framewOrk for Performance Exploration) with performance models for disk data read, cache performance, and page fault penalty. Second, an analytical performance model is also developed. We have evaluated our performance prediction framework as well as the analytical model on three hardware setups with well-known data mining algorithms implemented in three programming paradigms, MapReduce, MATE (a MapReduce-like parallel system with an alternate API for multi-core environments) and Smart (a MapReduce-like framework for in-situ analytics). Results show that our performance prediction framework along with the incorporated performance models are capable of accurately predicting execution times for parallel scientific analytics on different hardware setups.

97 MATHEMATICS AND COMPUTING↗

Searching for pulsars using image pattern recognition

In the modern era of big data, many fields of astronomy are generating huge volumes of data, the analysis of which can sometimes be the limiting factor in research. Fortunately, computer scientists have developed powerful data-mining techniques that can be applied to various fields. In this paper, we present a novel artificial intelligence (AI) program that identifies pulsars from recent surveys by using image pattern recognition with deep neural nets—the PICS (Pulsar Image-based Classification System) AI. The AI mimics human experts and distinguishes pulsars from noise and interference by looking for patterns from candidate plots. Different from other pulsar selection programs that search for expected patterns, the PICS AI is taught the salient features of different pulsars from a set of human-labeled candidates through machine learning. The training candidates are collected from the Pulsar Arecibo L-band Feed Array (PALFA) survey. The information from each pulsar candidate is synthesized in four diagnostic plots, which consist of image data with up to thousands of pixels. The AI takes these data from each candidate as its input and uses thousands of such candidates to train its ∼9000 neurons. The deep neural networks in this AI system grant it superior ability to recognize various types of pulsars as well as their harmonic signals. The trained AI's performance has been validated with a large set of candidates from a different pulsar survey, the Green Bank North Celestial Cap survey. In this completely independent test, the PICS ranked 264 out of 277 pulsar-related candidates, including all 56 previously known pulsars and 208 of their harmonics, in the top 961 (1%) of 90,008 test candidates, missing only 13 harmonics. The first non-pulsar candidate appears at rank 187, following 45 pulsars and 141 harmonics. In other words, 100% of the pulsars were ranked in the top 1% of all candidates, while 80% were ranked higher than any noise or interference. The performance of this system can be improved over time as more training data are accumulated. This AI system has been integrated into the PALFA survey pipeline and has discovered six new pulsars to date.

79 ASTRONOMY AND ASTROPHYSICS↗

Health Management Applications for International Space Station

Traditional mission and vehicle management involves teams of highly trained specialists monitoring vehicle status and crew activities, responding rapidly to any anomalies encountered during operations. These teams work from the Mission Control Center and have access to engineering support teams with specialized expertise in International Space Station (ISS) subsystems. Integrated System Health Management (ISHM) applications can significantly augment these capabilities by providing enhanced monitoring, prognostic and diagnostic tools for critical decision support and mission management. The Intelligent Systems Division of NASA Ames Research Center is developing many prototype applications using model-based reasoning, data mining and simulation, working with Mission Control through the ISHM Testbed and Prototypes Project. This paper will briefly describe information technology that supports current mission management practice, and will extend this to a vision for future mission control workflow incorporating new ISHM applications. It will describe ISHM applications currently under development at NASA and will define technical approaches for implementing our vision of future human exploration mission management incorporating artificial intelligence and distributed web service architectures using specific examples. Several prototypes are under development, each highlighting a different computational approach. The ISStrider application allows in-depth analysis of Caution and Warning (C&W) events by correlating real-time telemetry with the logical fault trees used to define off-nominal events. The application uses live telemetry data and the Livingstone diagnostic inference engine to display the specific parameters and fault trees that generated the C&W event, allowing a flight controller to identify the root cause of the event from thousands of possibilities by simply navigating animated fault tree models on their workstation. SimStation models the functional power flow for the ISS Electrical Power System and can predict power balance for nominal and off-nominal conditions. SimStation uses realtime telemetry data to keep detailed computational physics models synchronized with actual ISS power system state. In the event of failure, the application can then rapidly diagnose root cause, predict future resource levels and even correlate technical documents relevant to the specific failure. These advanced computational models will allow better insight and more precise control of ISS subsystems, increasing safety margins by speeding up anomaly resolution and reducing,engineering team effort and cost. This technology will make operating ISS more efficient and is directly applicable to next-generation exploration missions and Crew Exploration Vehicles.

Alena, Richard↗

Impact of the International Space Station Research Results

The International Space Station (ISS) facilitates research that benefits human lives on Earth and serves as the primary testing ground for technology development to sustain life in the extreme environment of space. To date, investigators have published a wide range of ISS science results, from improved theories about the creation of stars to the outcome of data mining “omics” repositories of previously completed ISS investigations. Because of the unique microgravity environment of the ISS laboratory and the multidisciplinary and international nature of the research, analyzing ISS scientific impacts is an exceptional challenge. As a result, the ISS Program Science Forum (PSF), made up of senior science representatives across the ISS international partnership, uses various methods to describe the impacts of ISS research activities. For the most part, past papers written by PSF members to assess the overall ISS research impact have focused on exhibiting ISS research impact by quantifying ISS research output or its perceived benefits for humanity. This paper proposes a new assessment of ISS impact from the perspective of the end users’ needs. To that end, the authors use visualizations and metrics of scientific publication data to show the ISS research influence on traditional scientific fields, its global reach and the benefits to people across the globe.

Diallo, Ousmane N.↗

A KNOWLEDGE DISCOVERY STRATEGY FOR RELATING SEA SURFACE TEMPERATURES TO FREQUENCIES OF TROPICAL STORMS AND GENERATING PREDICTIONS OF HURRICANES UNDER 21ST-CENTURY GLOBAL WARMING SCENARIOS

The connections among greenhouse-gas emissions scenarios, global warming, and frequencies of hurricanes or tropical cyclones are among the least understood in climate science but among the most fiercely debated in the context of adaptation decisions or mitigation policies. Here we show that a knowledge discovery strategy, which leverages observations and climate model simulations, offers the promise of developing credible projections of tropical cyclones based on sea surface temperatures (SST) in a warming environment. While this study motivates the development of new methodologies in statistics and data mining, the ability to solve challenging climate science problems with innovative combinations of traditional and state-of-the-art methods is demonstrated. Here we develop new insights, albeit in a proof-of-concept sense, on the relationship between sea surface temperatures and hurricane frequencies, and generate the most likely projections with uncertainty bounds for storm counts in the 21st-century warming environment based in turn on the Intergovernmental Panel on Climate Change Special Report on Emissions Scenarios. Our preliminary insights point to the benefits that can be achieved for climate science and impacts analysis, as well as adaptation and mitigation policies, by a solution strategy that remains tailored to the climate domain and complements physics-based climate model simulations with a combination of existing and new computational and data science approaches.

54 ENVIRONMENTAL SCIENCES↗

COMPLEX NETWORKS IN CLIMATE SCIENCE: PROGRESS, OPPORTUNITIES AND CHALLENGES

Networks have been used to describe and model a wide range of complex systems, both natural as well as man-made. One particularly interesting application in the earth sciences is the use of complex networks to represent and study the global climate system. In this paper, we motivate this general approach, explain the basic methodology, report on the state of the art (including our contributions), and outline open questions and opportunities for future research. Datasets and systems that can be represented as interaction networks (or graphs), broadly defined as any collection of interrelated objects or entities, have received considerable attention both from a theoretical viewpoint as well as various application domains; examples include the analysis of social networks, chemical interactions between proteins, the behavior of financial markets, and many others. Recently, the study of complex networks - that is, networks which exhibit non-trivial topological properties - has permeated numerous fields and disciplines spanning the physical, social, and computational sciences. So why do networks enjoy such broad appeal? Briefly, it is their ability to serve at once as a data representation, as an analysis framework, and as a visualization tool. The analytic capabilities in particular are quite powerful, as networks can uncover structure and patterns at multiple scales, ranging from local properties to global phenomena, and thus help better understand the characteristics of complex systems. We focus on one particular application of networks in the earth sciences, namely, the construction and analysis of climate networks. Identifying and analyzing patterns in global climate is an important task of growing scientific, social, and political interest, with the goal of deepening our understanding of the complex processes underlying observed phenomena. To this end, we make the case that complex networks offer a compelling perspective for capturing the dynamics of the climate system. Moreover, the computational sciences - specifically data mining and machine learning - are able to contribute a valuable set of methods and tools ranging from pattern recognition to predictive models. Thus, in this paper we expand upon the general approach to climate networks and motivate a promising area of interdisciplinary research. Indeed, we believe that this marriage of analytic methods, computational tools and domain science has the long-term potential for a transformative impact on our understanding of the earth's climate system.

54 ENVIRONMENTAL SCIENCES↗

MO-D-BRF-01: Pediatric Treatment Planning II: The PENTEC Report On Normal Tissue Complications

With advances in multimodality therapy, childhood cancer cure rates approach 80%. However, both radiotherapy and chemotherapy may cause debilitating or even fatal ‘late effects’ that are critical to understand, mitigate, or prevent. QUANTEC identified the uncertainties relating to side-effects of adult treatments, but this is more complicated for children in whom a mosaic of tissues develops at different rates and temporal sequences. Childhood cancer survivors have long life expectancy and may develop treatmentinduced secondary cancers and severe organ/tissue injury decades after treatment. Collaborative long-term observational studies and clinical research programs for survivors of pediatric and adolescent cancer provide some dose-response data for follow-up periods exceeding 40 years. Data analysis is challenging due to the influence of both therapeutic and developmental variables. PENTEC is a group of radiation oncologists, pediatric oncologists, subsepcialty physicians, medical physicists, biomathematic modelers/statisticians, and epidemiologists charged with conducting a critical synthesis of existing literature aiming to: critically analyze radiation dose-volume effects on normal tissue tolerances as a function of age/development in pediatric cancer patients in order to inform treatment planning and improve outcomes for survivors; describe relevant physics issues specific to pediatric radiotherapy; propose dose-volumeoutcome reporting standards to improve the knowledge base to inform future treatment guidelines. PENTEC has developed guidelines for systematic literature reviews, data extraction tolls and data analysis. This education session will discuss:1. Special considerations for normal tissue radiation response of children/adolescents, e.g. the interplay between development and radiotherapy effects.2. Epidemiology of organ/tissue injuries and secondary cancers.3. Exploration of dose-response differences between children and adults4. Methodology for literature review, data mining of outcomes databases, and NTCP or longitudinal modeling of doseresponse. 5. PENTEC goals and timetable. Learning Objectives: Understand important differences between normal tissue effects of radiation therapy in pediatric and adult patients. Be able to identify situations where there is ‘interplay’ between organ development and radiation-induced complications. Identify methods to systematically extract quantitative dose-volumeresponse relationships from existing outcomes databases. Provide guidance for the medical physicist to properly understand, implement, guide and control contemporary technology and applications in pediatric radiation oncology.

60 APPLIED LIFE SCIENCES↗

MAC: Memory Access Coalescer for 3D-Stacked Memory

Emerging data-intensive applications, such as graph analytics and data mining, exhibit irregular memory access patterns. Research has shown that with these memory-bound applications, traditional cache-based processor architectures, which exploit locality and regular patterns to mitigate the memory-wall issue, are inefficient. Meantime, novel 3D-stacked memory devices, such as Hybrid Mem- ory Cube (HMC) and High Bandwidth Memory (HBM), promise significant increases in bandwidth that appear extremely appealing for memory-bound applications. However, conventional memory interfaces designed for cache-based architectures and JEDEC DDR devices fit poorly with the 3D-stacked memory, which leads to significant under-utilization of the promised high bandwidth. As a response to these issues, in this paper we propose MAC (Memory Access Coalescer), a coalescing unit for the 3D-stacked memory. We discuss the design and implementation of MAC, in the context of a custom designed cache-less architecture targeted at data-intensive, irregular applications. Through a custom simulation infrastructure based on the RISC-V toolchain, we show that MAC achieves a coalescing efficiency of 52.85% on average. It improves the performance of the memory system by 60.73% on average for a large set of irregular workloads.

Wang, Xi↗

Fast and Accurate Support Vector Machines on Large Scale Systems

Support Vector Machines (SVM) is a supervised Machine Learning and Data Mining (MLDM) algorithm, which has become ubiquitous largely due to its high accuracy and obliviousness to dimensionality. The objective of SVM is to find an optimal boundary --- also known as hyperplane --- which separates the samples (examples in a dataset) of different classes by a maximum margin. Usually, very few samples contribute to the definition of the boundary. However, existing parallel algorithms use the entire dataset for finding the boundary, which is sub-optimal for performance reasons. In this paper, we propose a novel distributed memory algorithm to eliminate the samples which do not contribute to the boundary definition in SVM. We propose several heuristics, which range from early (aggressive) to late (conservative) elimination of the samples, such that the overall time for generating the boundary is reduced considerably. In a few cases, a sample may be eliminated (shrunk) pre-emptively --- potentially resulting in an incorrect boundary. We propose a scalable approach to synchronize the necessary data structures such that the proposed algorithm maintains its accuracy. We consider the necessary trade-offs of single/multiple synchronization using in-depth time-space complexity analysis. We implement the proposed algorithm using MPI and compare it with libsvm--- de facto sequential SVM software --- which we enhance with OpenMP for multi-core/many-core parallelism. Our proposed approach shows excellent efficiency using up to 4096 processes on several large datasets such as UCI HIGGS Boson dataset and Offending URL dataset.

Vishnu, Abhinav↗

SU-D-204-01: A Methodology Based On Machine Learning and Quantum Clustering to Predict Lung SBRT Dosimetric Endpoints From Patient Specific Anatomic Features

Purpose: To develop a data-mining methodology based on quantum clustering and machine learning to predict expected dosimetric endpoints for lung SBRT applications based on patient-specific anatomic features. Methods: Ninety-three patients who received lung SBRT at our clinic from 2011–2013 were retrospectively identified. Planning information was acquired for each patient, from which various features were extracted using in-house semi-automatic software. Anatomic features included tumor-to-OAR distances, tumor location, total-lung-volume, GTV and ITV. Dosimetric endpoints were adopted from RTOG-0195 recommendations, and consisted of various OAR-specific partial-volume doses and maximum point-doses. First, PCA analysis and unsupervised quantum-clustering was used to explore the feature-space to identify potentially strong classifiers. Secondly, a multi-class logistic regression algorithm was developed and trained to predict dose-volume endpoints based on patient-specific anatomic features. Classes were defined by discretizing the dose-volume data, and the feature-space was zero-mean normalized. Fitting parameters were determined by minimizing a regularized cost function, and optimization was performed via gradient descent. As a pilot study, the model was tested on two esophageal dosimetric planning endpoints (maximum point-dose, dose-to-5cc), and its generalizability was evaluated with leave-one-out cross-validation. Results: Quantum-Clustering demonstrated a strong separation of feature-space at 15Gy across the first-and-second Principle Components of the data when the dosimetric endpoints were retrospectively identified. Maximum point dose prediction to the esophagus demonstrated a cross-validation accuracy of 87%, and the maximum dose to 5cc demonstrated a respective value of 79%. The largest optimized weighting factor was placed on GTV-to-esophagus distance (a factor of 10 greater than the second largest weighting factor), indicating an intuitively strong correlation between this feature and both endpoints. Conclusion: This pilot study shows that it is feasible to predict dose-volume endpoints based on patient-specific anatomic features. The developed methodology can potentially help to identify patients at risk for higher OAR doses, thus improving the efficiency of treatment planning. R01-184173.

60 APPLIED LIFE SCIENCES↗

Scientific data analysis on data-parallel platforms.

As scientific computing users migrate to petaflop platforms that promise to generate multi-terabyte datasets, there is a growing need in the community to be able to embed sophisticated analysis algorithms in the computing platforms' storage systems. Data Warehouse Appliances (DWAs) are attractive for this work, due to their ability to store and process massive datasets efficiently. While DWAs have been utilized effectively in data-mining and informatics applications, they remain largely unproven in scientific workloads. In this paper we present our experiences in adapting two mesh analysis algorithms to function on five different DWA architectures: two Netezza database appliances, an XtremeData dbX database, a LexisNexis DAS, and multiple Hadoop MapReduce clusters. The main contribution of this work is insight into the differences between these DWAs from a user's perspective. In addition, we present performance measurements for ten DWA systems to help understand the impact of different architectural trade-offs in these systems.

99 GENERAL AND MISCELLANEOUS↗