Systematic Framework for Selecting Trajectories in Data Augmentation Shows Mixed Results Across Datasets
A new master's thesis published on arXiv introduces a framework comparing five strategies for selecting which trajectories to augment in machine learning datasets. The work tests Outlierness, Diversity, Representativeness, Uncertainty, and Random selection across four datasets spanning animal, maritime, and urban movement data. The findings suggest systematic selection can outperform random baselines in sparse datasets but may introduce noise in dense, high-quality datasets.
Researchers have long used trajectory data augmentation to address data scarcity in machine learning, but selecting which trajectories to augment has typically relied on naive random sampling. This thesis by Adam Nordlling systematically evaluates five selection strategies—Outlierness, Diversity, Representativeness, Uncertainty, and Random—across datasets covering fox behavior, deer behavior (Starkey), maritime AIS traffic, and urban car traffic. An Optuna-based hyperparameter optimization loop was integrated to identify the best augmentation parameters for each dataset. Results show that Outlierness and Uncertainty strategies offered greater stability and were less prone to performance degradation than random selection, particularly in sparse datasets. However, UMAP visualizations revealed that systematic augmentation can repair topological fragmentation in sparse data while acting as corrupting noise in dense datasets. The study also identified physical limitations in high-velocity domains, where standard perturbation techniques cause divergence in feature space. Overall, the work concludes that the value of trajectory augmentation is strictly conditional on dataset characteristics.
What's missing
The thesis is a master's project and has not undergone peer review. Key limitations include the restricted search space explored by the Optuna hyperparameter optimization, the generalizability of findings beyond the four tested datasets, and the lack of comparison against non-geometric augmentation methods. The extent to which results transfer to other trajectory domains (e.g., pedestrian or aerial data) remains an open question.
What different sources said
- arXiv cs.LGCenter
A Systematic Approach for Selecting Trajectories for Data Augmentation
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.