MétaCan
Menu
Back to cohort
Record W4297999823 · doi:10.1145/3563950

Synthetic Behavior Sequence Generation Using Generative Adversarial Networks

2022· article· en· W4297999823 on OpenAlex
Fateme Akbari, Kamran Sartipi, Norm Archer

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueACM Transactions on Computing for Healthcare · 2022
Typearticle
Languageen
FieldComputer Science
TopicContext-Aware Activity Recognition Systems
Canadian institutionsMcMaster University
Fundersnot available
KeywordsComputer scienceSimilarity (geometry)Generative grammarSequence (biology)Machine learningAdversarial systemReinforcement learningArtificial intelligenceVariety (cybernetics)Function (biology)Data miningPopulation

Abstract

fetched live from OpenAlex

Due to the increase in life expectancy in advanced societies leading to an increase in population age, data-driven systems are receiving more attention to support the older people by monitoring their health. Intelligent sensor networks provide the ability to monitor their activities without interfering with routine life. Data collected from smart homes can be used in a variety of data-driven analyses, including behavior prediction. Due to privacy concerns and the cost and time required to collect data, synthetic data generation methods have been considered seriously by the research community. In this article, we introduce a new Generative Adversarial Network (GAN) algorithm, namely, BehavGAN , that applies GAN to the problem of behavior sequence generation. This is achieved by learning the features of a target dataset and utilizing a new application for GANs in the simulation of older people’s behaviors. We also propose an effective reward function for GAN back-propagation by incorporating n-gram-based similarity measures in the reinforcement mechanism. We evaluate our proposed algorithm by generating a dataset of human behavior sequences. Our results show that BehavGAN is more effective in generating behavior sequences compared to MLE, LeakGAN, and the original SeqGAN algorithms in terms of both similarity and diversity of generated data. Our proposed algorithm outperforms current state-of-the-art methods when it comes to generating behavior sequences consisting of limited-space sequence tokens.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Science and technology studies
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.876
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0030.000
Scholarly communication0.0000.000
Open science0.0010.000
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.159
GPT teacher head0.353
Teacher spread0.194 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it