Reproduce “Investigating Manifold Neighborhood size for Nonlinear Analysis of LIBS Amino Acid Spectra” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Piyush K. Sharma, Gary Holness, Poopalasingam Sivakumar, Yuri Markushin, Noureddine Melikechi
Nonlinear manifold learning techniques, ISOMAP and LLE, were applied to Laser-Induced Breakdown Spectroscopy (LIBS) data of amino acids, identifying an optimal neighborhood size range for robust spectral analysis. This approach demonstrated superior classification accuracy for amino acid spectra compared to traditional linear methods.
Key insights
- An optimal neighborhood parameter interval of [8, 15] was identified for nonlinear manifold learning, balancing the capture of intra-sample differences and inter-sample separability.
- ISOMAP demonstrated high efficiency in variance explanation, capturing 98.8% to 99.93% of data variance with a single component, indicating strong underlying structure in LIBS spectra.
- The analysis revealed experimental artifacts, specifically brass alloy contributions from micro-well containers, highlighting the value of advanced data analysis for experimental quality control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.