Publications by authors named "John Cava"
LossLens: Diagnostics for Machine Learning Through Loss Landscape Visual Analytics.
IEEE Comput Graph Appl
December 2024
Modern machine learning often relies on optimizing a neural network's parameters using a loss function to learn complex features. Beyond training, examining the loss function with respect to a network's parameters (i.e.
View Article and Find Full Text PDFArticle Synopsis
- Researchers developed a metabolomics protocol for Lake Taihu, identifying the minimum biomass required for analyses and the times when metabolomes are stable, with phytoplankton requiring 475-950 µg DW and whole-water samples needing 940-1760 µg DW.
- The study found that phytoplankton (F) and whole-water (W) microbiomes showed asynchronous changes in taxonomical composition and metabolic states throughout the day, with F peaking in metabolic activity between 12-17 pm and W around 12 pm, revealing a significant overlap of 482 metabolites.
- A deeper examination highlighted differences in metabolic processes, such as carbon metabolism in F and amino acid metabolism in W, along with insights into genetic diversity and interactions
Predictive modeling of macromolecular recognition and protein-protein complementarity represents one of the cornerstones of biophysical sciences. However, such models are often hindered by the combinatorial complexity of interactions at the molecular interfaces. Exemplary of this problem is peptide presentation by the highly polymorphic major histocompatibility complex class I (MHC-I) molecule, a principal component of immune recognition.
View Article and Find Full Text PDF