2019
DOI: 10.1101/597518
|View full text |Cite
Preprint
|
Sign up to set email alerts
|

Peax Interactive Visual Pattern Search in Sequential Data Using Unsupervised Deep Representation Learning

Abstract: We present PEAX, a novel feature-based technique for interactive visual pattern search in sequential data, like time series or data mapped to a genome sequence. Visually searching for patterns by similarity is often challenging because of the large search space, the visual complexity of patterns, and the user's perception of similarity. For example, in genomics, researchers try to link patterns in multivariate sequential data to cellular or pathogenic processes, but a lack of ground truth and high variance mak… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
30
0

Year Published

2020
2020
2022
2022

Publication Types

Select...
4
1
1

Relationship

1
5

Authors

Journals

citations
Cited by 10 publications
(31 citation statements)
references
References 56 publications
1
30
0
Order By: Relevance
“…These typically present approaches where human involvement and trust are critical for utilization. As a result, several papers report bespoke, complex evaluation methodologies that go beyond more traditional pair analytics [KBJ∗20] and observational studies [CYL∗20; DSKE20; LPH∗20] that are also used.…”
Section: Application‐specific Evaluationsmentioning
confidence: 99%
See 2 more Smart Citations
“…These typically present approaches where human involvement and trust are critical for utilization. As a result, several papers report bespoke, complex evaluation methodologies that go beyond more traditional pair analytics [KBJ∗20] and observational studies [CYL∗20; DSKE20; LPH∗20] that are also used.…”
Section: Application‐specific Evaluationsmentioning
confidence: 99%
“…This dimension focuses on how the transparency of models was specifically communicated and evaluated in a system or study. This is a common focus of HCML papers, and often consists of showing the mechanisms of the models themselves (e.g., [KAS∗20; LPH∗20]). While early work in (X)AI equated model transparency with the presence of an explanation, later work found that transparency might be overwhelming [PGH∗21].…”
Section: Dimensions Of Analysismentioning
confidence: 99%
See 1 more Smart Citation
“…Users are not only able to query instances and label them via active learning, but also to understand and steer machine learning models interactively. This concept is also used in text document retrieval [25], sequential data retrieval [30], trajectory classification [27], identifying relevant tweets [37], and argumentation mining [38]. For example, to annotate text fragments in argumentation mining tasks, Sperrle et al [38] developed a language model for fragment recommendation.…”
Section: Label-level Improvementmentioning
confidence: 99%
“…We also see an opportunity to develop active learning or labeling pipelines developed in the machine learning community [47]. The visualization community traditionally focused on improving the labeling process for humans [4,5,22,28]. Perhaps it is time to take our own medicine and aid visualization evaluation by having algorithms learn which stimuli have to be evaluated by humans and which can be evaluated by machines.…”
Section: Opportunitiesmentioning
confidence: 99%