PARAFAC and missing values

By Giorgio Tomasi, Rasmus Bro. 2005

272 citations recorded by the source corpus.

Subfields: Analytical Chemistry, Electrochemistry, Industrial and Manufacturing Engineering

View Paper: https://www.semanticscholar.org/paper/a02d3a6275d337e0672446d3d7f20f9cc243abfa

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Analytical Chemistry

In Electrochemistry

In Industrial and Manufacturing Engineering

Citations in the corpus

Tensor Decompositions and Applications (2009, background): “Bro [28] mentions that PARATUCK2 is appropriate for problems that involve interactions between factors or when more ﬂexibility than CP is needed but not as much as Tucker.”
A scalable optimization approach for fitting canonical tensor decompositions (2011, methodology): “Thus, Tomasi and Bro [30,28] have argued that it is preferable to work instead with J T J using a Levenberg–Marquardt (LM) method.”
missMDA: A Package for Handling Missing Values in Multivariate Data Analysis (2016, methodology): “However, we note that related algorithms have been suggested for other multi-table and multi-way methods, including PARAFAC (Tomasi and Bron 2005; Kroonenberg 2008; Acar, Dunlavy, Kolda, and Mrup 2011; Smilde, Bro, and Geladi 2004).”
Unsupervised Multiway Data Analysis: A Literature Survey (2009, methodology): “Tomasi and Bro [57] highlight the advantages of SWATLD algorithm in the case of overfactoring and also suggest PMF3 [58] and Levenberg-Marquadt algorithm [58], [59] as alternative algorithms with better convergence properties than ALS for fitting a PARAFAC model.”
A comparison of algorithms for fitting the PARAFAC model (2006, methodology): “The second set is represented by PMF3 (Positive Matrix Factorisation for 3-way arrays (Paatero, 1997)) and dGN (damped Gauss–Newton, also known as Levenberg–Marquadt (Paatero, 1997; Tomasi and Bro, 2004)).”
Tensors for Data Mining and Data Fusion (2016, methodology): “The works of Tomasi and Bro [2005] and Acar et al. [2010, 2011] show how this can be done algorithmically.”
Scalable Tensor Factorizations for Incomplete Data (2010, methodology): “For instance, [7] shows that, in spectroscopy, intermittent machine failures or different sampling frequencies may result in tensors with missing fibers (i.”
On Tensors, Sparsity, and Nonnegative Factorizations (2011, unknown): “One may also consider extensions of this work in the context of missing data [22, 7, 50, 1] and for alternative tensor factorizations such as Tucker [17].”
Low Complexity Damped Gauss-Newton Algorithms for CANDECOMP/PARAFAC (2012, unknown): “The Jacobian matrixJ has a form of [20,31]”