1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 10111010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 10111010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
AIM is developing new analysis algorithms to provide continuous, automated synthesis of new knowledge and to enable
measurement systems to be steered in response to emerging knowledge, rebalancing the effort between humans and machines.
APPROACH
USE CASES
Interaction with human interfaces to implicitly
weight, tune, and modify underlying models
Integration framework and testing range
Foundational streaming algorithms, methods for extracting features from
streams
Scalable symbolic deduction and incremental machine learning to track a stream
Human-Machine Feedback
Hypothesis Generation & Testing
Streaming Data Characterization & Processing
Work Environments
Instrumentation to measure overall accuracy,
utility, and throughput
Data reduction techniques for higher effective throughput
Generate, update, and validate human-
understandable hypotheses from streaming classifiers
Visual strategies for bidirectional communication
of and interaction with multiple hypotheses
In a world where data are continually streaming from distributed and diverse sources—from scientific instruments, to web traffic, to live imagery—making timely discoveries requires computing capabilities that can keep pace with rapidly evolving phenomena.
Analysis in Motion InitiativeInteractive Streaming Analytics at Scale
Imaging of Dynamic Processes in Electron Microscopy
Electron microscopes are used in the biological and material sciences,
however challenges are faced during data acquisition. Compressed
sensing addresses these challenges; AIM’s research focuses on
enabling users to interact with the compressed image streams.
Insider Threat Detection in the Cloud
With the migration of global computing environments to commercial cloud computing solutions, new cybersecurity challenges are being introduced. AIM is tackling this challenge head-on by developing novel, scientifically-rigorous approaches and methodologies, specifically as applied to insider threat detection.
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1011 0101 0010 1110 1011 1010 1010 1110 1111 0100 0101 0101 0101 1010 1011
1010 1110 1111 0100 0101 0101 0101 1010 1011 1010 1011 0101 0010 1110 1011 1010
Collaborate with us
We're interested in partnering with
individuals and organizations. If you have
expertise in the areas of machine learning
or human-computer interaction, and are
interested in partnership opportunities,
then we'd like to talk with you.
AIM by the numbers
Initiative Lead: Mark Greaves • [email protected] • (206) 528-3300
THE R&D AGENDA
1123conference
papers27
abstracts
formalreports
journal articles
25 university subcontracts
40 seminar speakers hosted
9 invention disclosures168 staff / interns
How AIM is sharing its research
June 2017 PNNL SA XXXXX
InsightsStreaming Data Characterization
Stream Processing Algorithms
User ExperienceData Stream
Modeling Continuous
Human Information
Processing (MCHIP)Leslie Blaha
Streaming Data
Characterization (SDC)Mark Greaves
Streaming Query User
Interface (SQUINT)Svitlana Volkova
TranspireAritra Dasgupta
Stream Adaptive Foraging
for Evidence (SAFE)Nicole Nichols
AIM Streaming
Infrastructure (ASI)Dimitri Zarzhitsky
Bounded Informational Framework
for the Optimization of Streaming
systems (BIFROST)Luke Gosink
Dynamic Network Analysis
via Motifs (DYNAMO)George Chin
Temporal Modeling in
Streaming Analytics
(TeMpSA)Lisa Bramer
11
AIM is developing an exploratory process of defining new knowledge in a streaming environment; we seek to construct machines that can help tell stories from data by constructing and evolving candidate hypotheses in parallel with the stream.
85presentations
Strategic Surprise
Computing advances as applied to the nuclear trafficking domain
Evolution of Nuclear Magnetic Resonance (NMR)
Enabling faster, more accurate NMR metabolomics
PREVIOUS USE CASES