Open Lab
Explore chemometrics and analytical science through theory, methods, code, and practice.
34 resources
34 resources
ComDim
Common Components and Specific Weights Analysis
Unsupervised CCSWA / ComDim: common sample-space dimensions with component-specific block saliences, not PCA of a concatenated superblock.
LW-PLS
Local Partial Least Squares Regression
Query-specific KNN-LWPLSR: select nearest calibration neighbors, assign rchemo-style observation weights, and fit weighted PLS regression for each new sample.
LW-PLS-DA
Locally Weighted Partial Least Squares Discriminant Analysis
Query-specific local PLS-DA: select n_local neighbors, apply Bevilacqua and Marini Table 1 distance weights through X_w = W X_l, and assign the class of the largest predicted dummy component.
MULTIBLOCK
Multiblock Data Analysis & Data Fusion
A conceptual map of linked data blocks, fusion levels, common and distinct subspaces, and the algebraic choices that distinguish chemometric multiblock methods.
MB-PCA
Multiblock PCA / Consensus PCA
Unsupervised multiblock PCA in the CPCA-W / SUM-PCA equivalent formulation: processed superblock PCA with block-level contributions, loadings, and scores.
MB-PLS
Multiblock Partial Least Squares
Supervised multiblock PLS with multiple predictor blocks and one response block, in the Westerhuis 1998 super-score-deflation formulation equivalent to standard PLS on a sqrt(J_b)-weighted superblock.
PARAFAC
Parallel Factor Analysis
A three-way CANDECOMP/PARAFAC decomposition that models a tensor as a sum of rank-one outer products and estimates all factor matrices jointly by alternating least squares.
SO-PLS
Sequential and Orthogonalized Partial Least Squares
Supervised sequential multiblock PLS: each later predictor block is orthogonalized to previously extracted PLS scores, then used to model the current response residual.
HCA
Hierarchical Cluster Analysis
Unsupervised agglomerative clustering that starts from singleton observations and merges clusters under a chosen distance metric and linkage criterion, producing a nested hierarchy.