Visual discovery and model-driven explanation of time series patterns

@article{Sarkar2016VisualDA,
  title={Visual discovery and model-driven explanation of time series patterns},
  author={Advait Sarkar and Martin Spott and Alan F. Blackwell and Mateja Jamnik},
  journal={2016 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)},
  year={2016},
  pages={78-86}
}
  • Advait SarkarM. Spott M. Jamnik
  • Published 1 September 2016
  • Computer Science
  • 2016 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC)
Gatherminer is an interactive visual tool for analysing time series data with two key strengths. First, it facilitates bottom-up analysis, i.e., the detection of trends and patterns whose shapes are not known beforehand. Second, it integrates data mining algorithms to explain such patterns in terms of the time series' metadata attributes - an extremely difficult task if the space of attribute-value combinations is large. To accomplish these aims, Gatherminer automatically rearranges the data to… 

Figures from this paper

MTV: Visual Analytics for Detecting, Investigating, and Annotating Anomalies in Multivariate Time Series

This paper investigates current practices used to detect and investigate anomalies in time series data in industrial contexts and identifies corresponding needs and introduces MTV, a visual analytics system to support such workflow.

Assisted pattern mining for discovering interactive behaviours on the web

Towards User‐Centered Active Learning Algorithms

This work identified computational building blocks of user strategies, formalize them, and investigate their potentials for different machine learning tasks in systematic experiments, and observed that data‐based user strategies work considerably well in early phases, while model‐ based user strategies perform better during later phases.

Visual-Interactive Semi-Supervised Labeling of Human Motion Capture Data

A visual-interactive approach for the semi-supervised labeling of human motion capture data that combines multiple views supporting the user in the visualinteractive labeling process and guided selection of candidate data for the extension and improvement of the labeling closes the feedback loop of the semisupervised workflow.

Perceptual Visual Interactive Learning

A perceptual visual interactive learning (PVIL) framework is proposed, which adopts gestalt principle to design interaction strategy and multi-dimensionality reduction (MDR) to optimize the process of visualization.

IRVINE: A Design Study on Analyzing Correlation Patterns of Electrical Engines

The results suggest a high usability and usefulness of IRVINE as part of the improvement of a real-world manufacturing process, and domain experts were able to label and annotate produced electrical engines more than 30% faster.

Comparing Visual-Interactive Labeling with Active Learning: An Experimental Study

It is found that visual-interactive labeling can outperform active learning, given the condition that dimension reduction separates well the class distributions and using dimension reduction in combination with additional visual encodings that expose the internal state of the learning model turns out to improve the performance of visual-Interactive labeling.

VIAL: a unified process for visual interactive labeling

This paper studies the process of labeling data instances with the user in the loop, from both the machine learning and visual interactive perspective, and proposes the “visual interactive labeling” (VIAL) process that unifies both approaches.

Difference based query strategies in active learning

Two active learning methods are proposed in the machine learning literature based on difference calculation idea and both of the difference based sampling query strategies surpass the competitive methods in literature.

2022

References

SHOWING 1-10 OF 39 REFERENCES

Visual exploration of frequent patterns in multivariate time series

Four novel visual analytics methods are introduced to interactively examine motifs and gain new insights into the recurring patterns to analyze system operations, and both power consumption and server utilization in data centers are predicted.

Visualizing and Discovering Non-Trivial Patterns in Large Time Series Databases

VizTree is a time series pattern discovery and visualization system based on augmenting suffix trees that provides novel interactive solutions to many pattern discovery problems, including the discovery of frequently occurring patterns (motif discovery), surprising patterns (anomaly detection), and query by content.

Interactive pattern search in time series

TimeSearcher 2, a tool that enables users to explore multidimensional data using synchronized tables and graphs with overview+detail, filter the time series data to reduce the scope of the search, select an existing pattern to find similar occurrences, and interactively adjust similarity parameters to narrow the result set is presented.

Guided discovery of interesting relationships between time series clusters and metadata properties

A novel approach for guiding users in discovering interesting relationships between clusters and associated metadata, and presents a system implementing an encompassing, yet extensible, set of interestingness scores for categorical metadata, which can be extended to numerical metadata.

Multi-Resolution Techniques for Visual Exploration of Large Time-Series Data

A framework for intelligent time- and data-dependent visual aggregation of data along multiple resolution levels is proposed, which leads to effective visualization support for long time-series data providing both focus and context.

VISUALIZATION METHODS FOR TIME-DEPENDENT DATA-AN OVERVIEW

An overview on the visualization of time-series data and the available techniques is given, a taxonomy is provided and general aspects ofTime-dependent data is discussed, and techniques for analysing time-dependent multivariate data sets are discussed in more detail.

Observation-level interaction with statistical models for visual analytics

This paper explores two possible observation-level interactions, namely exploratory and expressive, within the context of three statistical methods, Probabilistic Principal Component Analysis (PPCA), Multidimensional Scaling (MDS), and Generative Topographic Mapping (GTM).

VisInfo: a digital library system for time series research data based on exploratory search—a user-centered design approach

This work presents VisInfo, a web-based digital library system (DLS) with the goal to provide visual access to time series research data, and suggests a generalized project workflow that may be applied in related, prospective approaches.

Feedback-driven interactive exploration of large multidimensional data supported by visual classifier

This work introduces a framework for a feedback-driven view exploration, inspired by relevance feedback approaches used in Information Retrieval, and presents an instantiation of the framework for exploration of Scatter Plot Spaces based on visual features.

Dis-function: Learning distance functions interactively

It is illustrated empirically that with only a few iterations of interaction and optimization, a user can achieve a scatterplot view and its corresponding distance function that reflect the user's knowledge of the data.