@INPROCEEDINGS{lu10,
  author = {Lu, Z. and Savas, B. and Tang, W. and Dhillon, I. S.},
  title = {Link prediction using multiple sources of information},
  booktitle = {Proceedings of the IEEE International Conference on Data Mining (ICDM)},
  year = {2010},
  pages = {923--928},
}
@TECHREPORT{lu10tr,
  author = {Lu, Z. and Savas, B. and Tang, W. and Dhillon, I. S.},
  title = {Link prediction using multiple sources of information},
  institution = {Department of Computer Science, The University of Texas at Austin},
  year = {2010},
  number = {TR--10--35}
}

@ARTICLE{sael10,
  author = {Savas, B. and Eld\'{e}n, L.},
  title = {{K}rylov-Type Methods for Tensor Computations},
  journal = {Submitted to Linear Algebra and its Applications},
  year = {2010}
}
@article{savaslim10,
author = {Savas, B. and Lim, L.-H.},
title = {{Q}uasi-{N}ewton Methods on {G}rassmannians and Multilinear Approximations of Tensors},
publisher = {SIAM},
year = {2010},
journal = {SIAM Journal on Scientific Computing},
volume = {32},
number = {6},
pages = {3352--3393},
keywords = {Grassmann manifold; Grassmannian; product of Grassmannians; Grassmann quasi-Newton; Grassmann BFGS; Grassmann limited memory BFGS; multilinear rank; symmetric multilinear rank; tensor; symmetric tensor; approximations},
url = {http://link.aip.org/link/?SCE/32/3352/1},
doi = {10.1137/090763172}
}   
@TECHREPORT{sael:09,
  author = {Berkant Savas and Lars Eld\'{e}n},
  title = {Krylov subspace methods for tensor computations},
  institution = {Department of Mathematics, Linköpings Universitet},
  year = {2009},
  number = {LITH-MAT-R-2009-02-SE},
}

@TECHREPORT{Savas:08,
  author = {Berkant Savas and Lek-Heng Lim},
  title = {Best multilinear rank approximation of tensors with quasi-Newton methods on Grassmannians},
  institution = {Department of Mathematics, Linköpings Universitet},
  year = {2008},
  number = {LITH-MAT-R-2008-01-SE},
}
@ARTICLE{elsa09,
  author = {L. Eld\'en and B. Savas},
  title = {A {Newton--Grassmann} method for computing the Best Multi-Linear
    Rank-(${r}_1,r_2,r_3$) Approximation of a Tensor},
  journal = {SIAM J. Matrix Anal. Appl.},
  year = {2009},
  volume = {31},
  pages = {248-271},
  abstract = {We derive a Newton method for computing the best rank-$(r_1,r_2,r_3)$
	approximation of a given $J \x K \x L$ tensor $\cA$. The problem
	is formulated as an approximation problem on a product of Grassmann
	manifolds. Incorporating the manifold structure into Newton's method
	ensures that all iterates generated by the algorithm are points on
	the Grassmann manifolds. We also introduce a consistent notation
	for matricizing a tensor, for contracted tensor products and some
	tensor-algebraic manipulations, which simplify the derivation of
	the Newton equations and enable straightforward algorithmic implementation.
	Experiments show a quadratic convergence rate for the Newton-Grassmann
	algorithm.}
} 
@ARTICLE{Savas:07,
  author = {Savas, Berkant and Eldén, Lars},
  title = {Handwritten digit classification using higher order singular value
	decomposition},
  journal = {Pattern Recognition},
  year = {2007},
  volume = {40},
  pages = {993 – 1003},
  abstract = {In this paper we present two algorithms for handwritten digit classification
	based on the higher order singular value decomposition (HOSVD).
	The first algorithm uses HOSVD for construction of the class models
	and achieves classification results with error rate lower than 6%.
	The second algorithm uses the HOSVD for tensor approximation simultaneously
	in two modes. Classification results for the second algorithm are
	almost down at 5% even though the approximation reduces the original
	training data with more than 98% before the construction of the
	class models. The actual classification in the test phase for both
	algorithms is conducted by solving a series least squares problems.
	Considering computational amount for the test presented the second
	algorithm is twice as efficient as the first one.},
  doi = {10.1016/j.patcog.2006.08.004}
}
@ARTICLE{Savas:06b,
  author = {Savas, Berkant and Lindgren, David},
  title = {Rank reduction and volume minimization approach to state-space subspace
	system identification},
  journal = {Signal processing},
  year = {2006},
  volume = {86},
  pages = {3275–3285},
  doi = {10.1016/j.sigpro.2006.01.008}
}
@ARTICLE{Savas:06a,
  author = {Berkant Savas},
  title = {Dimensionality reduction and volume minimization — generalization
	of the determinant minimization criterion for reduced rank regression
	problems},
  journal = {Linear Algebra and its Applications},
  year = {2006},
  volume = {418},
  pages = {201-214},
  doi = {10.1016/j.laa.2006.01.032}
}
@ARTICLE{Elden:05,
  author = {Eldén, Lars and Savas, Berkant},
  title = {The maximum likelihood estimate in reduced-rank regression},
  journal = {Numerical Linear Algebra with Applications},
  year = {2005},
  volume = {12},
  pages = {731-741},
  doi = {10.1002/nla.447}
}