Search engine run on: http://users.monash.edu.au/
Glookbib search for: zz0321
%A A. K. Lenstra
%A H. W. Lenstra jr.
%A L. Lovasz
%T Factoring polynomials with rational coefficients
%J Mathematische Annalen
%V 261
%P 515-534
%D 1982
%K jrnl, c1982, c198x, c19xx, zz0321, LLL algorithm, polynomial, factoring,
factors, vector, lattice basis reduction, nearly orthogonal, orthogonalise
%X "... present a polynomial-time alg. to solve ...: given a non-zero polynomial
f in Q(X) in one variable with rational coeffs., find the decomposition of f
into irreducible factors in Q(X). It is well known that this is equiv. to
factoring primitive polynomials feZ[X] into irreducible factors in Z[X]. Here
we call f in Z[X] primitive if the GCD of its coeffs. (the content of f)
is 1. Our alg. performs well in practice, cf.(8). Its running time, measured
in bit operations, is O(n^{1/2}+n^9(log(|f|)^3)."
-- [doi:10.1007%2FBF01457454]['21].
Also see, "The Lenstra-Lenstra-Lovasz (LLL) lattice basis red'n alg. ...
calculates an LLL-reduced (short, nearly orthogonal) lattice basis in
time O(d^5 n log^3 B) ..."
-- LLL@[wikip]['21].
[Also search for: lattice basis] and [also search for: Gram Schmidt].
%A P. Bille
%A I. Li Gortz
%T The tree inclusion problem: In linear space and faster
%J ACM Trans. on Algorithms
%P #38
%M JUL
%D 2011
%K jrnl, c2011, c201x, c20xx, zz0321, tree inclusion, subtree,
problem, algorithm
%X "Given two rooted, ordered, & labeled trees P & T the tree inclusion problem
is to determine if P can be obtained from T by deleting nodes in T. ... has
recently been recognized as an important query primitive in XML databases.
Kilpelainen & Mannila ['95] presented the first polynomial-time alg. using
quadratic time & space. Since then several improved results have been
obtained for special cases when P & T have a small # of leaves or small
depth. However, in the worst case these algs. still use quadratic time &
space. Let n_S, l_S, & d_S denote the # of nodes, the # of leaves, & the
depth of a tree S in P, T. . . . show that the tree inc. problem can be
solved in space O(n_T) & time:
O(min(l_P n_T, l_P l_T log log n_T + n_T, (n_P n_T)/(log n_T + n_T log n_T)).
This improves or matches the best known time complexities while using only
linear space instead of quadratic. ... important in practical applications,
such as XML databases, where the space is likely to be a bottleneck."
-- [doi:10.1145/1978782.1978793]['21] and
0608124@[arXiv]['21].
%A Yuansi Chen
%T An almost constant lower bound of the isoperimetric coefficient in the
KLS conjecture
%J arXiv
%M NOV
%D 2020
%K TR, c2020, c202x, c20xx, zz0321, Burgain, KLS conjecture, convex,
high dimension, geometry
%T "We prove an almost constant lower bound of the isoperimetric coefficient in
the KLS conjecture. The lwb has the dimension dependency d-od(1). When the
dimension is large enough, our lwb is tighter than the prev. best bound which
has the dimension dependency d-1/4. Improving the current best lwb of the
isoperimetric coefficient in the KLS conjecture has many implications, inc.
improvements of the current best bounds in Bourgain's slicing conjecture & in
the thin-shell conjecture, better concentration inequalities for Lipschitz
fns of log-concave measures and better mixing time bounds for MCMC sampling
algs. on log-concave measures."
-- 2011.13661@[arXiv]['21].
(Also see news@[www]['21].)
%A J. Morris
%A Fang Song
%T Simple vertex coloring algorithms
%J arXiv
%M FEB
%D 2021
%K TR, c2021, c202x, c20xx, zz0321, maths, graph, vertex colouring algorithm
%X "Given a graph G with n vertices and maximum degree D, it is known that G
admits a vertex coloring with D+1 colors s.t. no edge of G is monochromatic.
This can be seen constructively by a simple greedy alg., which runs in time
O(n.D). Very recently, a seq. of results ... we give a simple alg. for
(1+eps)D-coloring. ..."
-- 2102.07089@[arXiv]['21].
%A L. Allison
%A A. S. Konagurthu
%A D. F. Schmidt
%T On universal codes for integers: Wallace Tree, Elias Omega and beyond
%J Data Compression Conference
%I IEEE
%P 313-322
%M MAR
%D 2021
%K conf, DCC, DCC21, DCC2021, c2021, c202x, c20xx, zz0321, universal code,
ints, integer, intCodes, Elias omega, omega2, omega*, omega star, Fibonacci,
Wallace Tree code, WTC, probability distribution, compression, entropy,
LAllison, ArunK
%X "A universal code for the (positive) integers is a variable length code that
can be used to store or compress a sequence of integers. It also implies a
probability distribution on integers which can be a natural choice when the
true distribution of a source of integers is unknown; such a code and
distribution may be useful in statistical inference. This paper provides two
improvements to the theory and practice of universal codes. First, it defines
and examines a new universal code omega* (omega-star) that asymptotically
beats the Elias omega code. Second, it analyses the properties of a code
proposed by Wallace based on trees, and shows it to be a universal code, to
have desirable properties for use in inference, and to beat the Elias omega
code on almost all integers up to the 1697-bit code-word mark. Encoding and
decoding routines for the codes described here are implemented and available
for interactive use[1]." Also see [www].
-- [more1],
[more2]['21] inc. source code,
[doi:10.1109/DCC50243.2021.00039]['22].
%A Swayambhoo Jain
%A et al
%T Low rank based end-to-end deep neural network compression
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, MolBio, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, NN, ANN, DNN,
compress, simplify
%X "Deep neural networks (DNNs), despite their performance on a wide variety of
tasks, are still out of reach for many applications as they require sig.
comp'l resources. ... we present a low-rank based end-to-end deep NN
compression frame-work with the goal of enabling DNNs performance to
computationally constrained devices.The proposed framework includes
techniques for low-rank based structural approx'n, quantization & lossless
arith.coding. Many of these techniques have been accepted in the MPEG working
draft on compressed NN Representations. We demonstrate the efficacy of the
proposed framework via extensive experiments on a variety of DNNs for various
tasks considered in this standardization activity. These techniques provide
impressive performance on DNNs used in ImageNet Large-Scale Visual
Recognition Challenge by compressing VGG16 by 61x, ResNet50 by almost 15x, &
MobileNetV2 by almost 7x.
-- [doi:10.1109/DCC50243.2021.00031]['21].
[Also search for: DCC21].
%A M. Soltani
%A et al
%T Compressing deep networks using Fisher score of feature maps
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, MolBio, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, NN, ANN, DNN,
deep, neural network, compress, simplify, map
%K conf, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, NN, ANN,
simplification, compress
%X "... propose a new structural technique for pruning deep neural networks with
skip-connections. ... based on measuring the importance of feature maps in
predicting the output of the model using their Fisher scores. These scores
subsequently used for removing the less informative layers from the graph of
the n/w. Extensive experiments on the classification of CIFAR-10, CIFAR-100,
& SVHN data sets demonstrate the efficacy of our compressing method both in
the # of parameters & operations.
-- [doi:10.1109/DCC50243.2021.00083]['21].
[Also search for: DCC21].
%A Y. Idelbayev
%A M. A. Carreira-Perpinan
%T Neural network compression via additive combination of reshaped, low-rank
matrices
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, NN, ANN,
simplification, compress
%X "... NN compression has become an important problem due to the increasing
necessity of running complex networks on small devices. We consider a form of
n/wk compression that has not been explored before: an additive combination
of reshaped low-rank matrices. That is, given the weights of a NN, we
constrain them as a sum of differently shaped low-rank matrices to reduce the
n/wk's size & inference demands. Computationally, this is a hard problem ...
We formulate it as a model selection over the family of compressed models &
give an optimization alg. that efficiently handles the inherent
combinatorial structure. This results in a 'Learning-Compression' alg. which
alternates between a std machine learning step & a step involving signal
compression. We demonstrate the effectiveness of the proposed compression
scheme & the corr. alg. on multiple networks and datasets."
-- [doi:10.1109/DCC50243.2021.00032]['21].
[Also search for: DCC21].
%A B. P. Blackburne
%A S. Whelan
%T Measuring the distance between multiple sequence alignments
%J Bioinformatics
%V 28
%N 4
%P 495-502
%M FEB
%D 2012
%K MolBio, c2012, c201x, c20xx, zz0321, distance between sequence alignments,
difference, multiple alignment, MSA, sum pairs, total column count, score
%X "... By measuring the differences between inferred alignments, we may be able
to develop an understanding of how these differences (i) relate to the
objective fns & heuristics used in MSA methods, & (ii) affect downstream
analyses. Results: We introduce four metrics to compare MSAs, which include
the posn in a seq. where a gap occurs or the location on a phylogenetic tree
where an insertion or deletion (indel) event occurs. ..."
-- [doi:10.1093/bioinformatics/btr701]['21].
%A R. Singh
%A et al
%T Unsupervised manifold alignment for single-cell multi-omics data
%J bioRxiv
%M JUN
%D 2020
%K TR, MolBio, c2020, c202x, c20xx, zz0321, MMD-MA, MMDMA, scRNAseq, scRNA,
manifold alignment, learning
%X "Integrating single-cell measurements that capture different properties of
the genome is vital to extending our understanding of genome biology. This
task is challenging due to the lack of a shared axis across datasets obtained
from different types of single-cell experiments. ... unsupervised algs. that
are capable of aligning single-cell experiments are critical to learning an
in silico co-assay that can help draw corr. among the cells. Maximum mean
discrepancy-based manifold alignment (MMD-MA) is such an unsupervised alg.
... we apply MMD-MA to two recent datasets that measure transcriptome &
chromatin accessibility in ~2000 single cells ..."
-- [doi:10.1101/2020.06.13.149195]['21],
[bioRxiv]['21].
Talk: [www]['20].
[Also search for: scRNAseq].
%A T. van der Valk
%A et al
%T Million-year-old DNA sheds light on the genomic history of mammoths
%J Nature
%M FEB
%D 2021
%K MolBio, c2021, c202x, c20xx, zz0321, ancient DNA, fossil, woolly mammoth,
M. primigenius, mamoth, Pleistocene
%X "... report the recovery of genome-wide data from three mammoth specimens
dating to the Early and Middle Pleistocene subepochs, two of which are more
than one million years old. We find that two distinct mammoth lineages were
present in eastern Siberia during the Early Pleistocene. ..."
-- [doi:10.1038/s41586-021-03224-9]['21].
[Also search for: MolBio ancient DNA].
%A D. Diaz-Dominguez
%A G. Navarro
%T A grammar compressor for collections of reads with applications to the
construction of the BWT
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, MolBio, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021,
DNA sequencing, grammar compressor, compression
%X "We describe a grammar for DNA sequencing reads from which we can compute the
BWT directly. Our motivation is to perform in succinct space genomic analyses
that require complex string queries not yet supported by repetition-based
self-indexes. Our approach is to store the set of reads as a grammar, but
when required, compute its BWT to carry out the analysis by using
self-indexes. Our experiments in real data showed that the space red'n we
achieve with our compressor is competitive with LZ-based methods & better
than entropy-based approaches. Compared to other popular grammars, in this
kind of data, we achieve, on avg., 12% extra compression & require less
working space and time."
-- [doi:10.1109/DCC50243.2021.00016]['21].
[Also search for: DCC21].
%A T. Freeth
%A et al
%T A model of the cosmos in the ancient Greek Antikythera mechanism
%J Scientific Reports
%V 11
%N #5821
%M MAR
%D 2021
%K jrnl, c2021, c202x, c20xx, zz0321, Antikythera mechanism, ancient, Greek,
astronomy, device, cosmos, history, computing, calculator
%X "The Antikythera Mechanism, an ancient Greek astronomical calculator, has
challenged researchers since its discovery in 1901. Now split into 82
fragments, only a third of the original survives, including 30 corroded
bronze gear wheels. Microfocus X-ray Computed Tomography (X-ray CT) in 2005
decoded the structure of the rear of the machine but the front remained
largely unresolved. X-ray CT also revealed inscriptions describing the
motions of the Sun, Moon and all five planets known in antiquity ..."
-- [doi:10.1038/s41598-021-84310-w].
(Also see [bbc]['21].)
[Also search for: Antikythera mechanism].
%A H. Else
%A R. Van Noorden
%T The fight against fake-paper factories that churn out sham science
%J Nature
%V 591
%P 516-519
%M MAR
%D 2021
%K news, c2021, c202x, c20xx, zz0321, papers, publications, paper mill, China,
PaperMill, scam, fraud, scientific research, cheating, ghostwriter,
academia, academic misconduct
%X "When Laura Fisher noticed striking similarities between research papers
submitted to RSC Advances, she grew suspicious. None of the papers had
authors or institutions in common, but their charts and titles looked
alarmingly similar, says Fisher, the executive editor at the journal.
'I was determined to try to get to the bottom of what was going on.' ..."
-- [doi:10.1038/d41586-021-00733-5]['21].
[Also search for: PaperMill].
%T Generating conjectures on fundamental constants with the Ramanujan Machine
%J Nature
%V 590
%P 67-73
%M FEB
%D 2021
%K jrnl, Ramanujan Machine, c2021, c202x, c20xx, zz0321, AI, maths, conjecture,
equations
%X "... Our algs. find dozens of well known formulas as well as previously
unknown ones, such as continued fraction representations of pi, e, Catalan's
constant, and values of the Riemann zeta fn. Several conjectures found by our
algs. were (in retrospect) simple to prove, whereas others remain as yet
unproved. ..."
-- [doi:10.1038/s41586-021-03229-4]['21].
Also see [ramanujanmachine]['21].
%A M. K. Varanasi
%A B. Aazhang
%T Parametric generalized Gaussian density estimation
%J J. of the Acoustical Soc. of America
%V 86
%P 1404
%D 1989
%K jrnl, c1989, c198x, c19xx, zz0321, generalized Gaussian density, GGD, normal,
exponential power, probability distribution, estimate, estimator, gnorm,
shape parameter, generalised
%X "... compare the large-sample as well as the small-sample properties of
different methods for estimating the parameters of a three-parameter
generalized Gaussian distribution. Three estimators, namely, the moment
method (MM), the maximum-likelihood (ML), & the moment/Newton-step (MNS)
estimators ..."
-- [doi:10.1121/1.398700]['21].
Also see,
"... is either of two families of parametric continuous prob. distn on the
real line. Both families add a shape parameter to the normal distn ...
(1) AKA exponential power distn, or the generalized error distn, this is a
parametric family of symmetric distns. It includes all normal & Laplace
distns & as limiting cases it includes all cts uniform distns on
bounded intervals of the real line. ...
(2) ... a family of continuous prob. distns in which the shape parameter can
be used to introduce skew. ..."
-- [wikip]['21].
(And see
gnorm@[r-project.org]['21].)
%A T. Ameen ur Rahman
%A et al ...
%T L. R. Varshney
%T The twelvefold way of non-sequential lossless compression
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, data sets, dataset,
twelvefold, twelve, unordered, equivalence, symmetry, compress
%X "... huge amounts of data that are not simple sequences of distinguishable
symbols, but have certain equivalence classes for patterns of symbols within
which lossless rep'n does not require unique indices. As a simple example,
the ordering of seqs. of scientific data records are often irrelevant & so
only the histogram or type class must be represented. Such rep'n problems
arise for biol. data, social graph structure, NN architecture, & elsewhere
due to the functional process generating or using the data. These
non-sequential info.sources may be studied implicitly via group theory & some
may explicitly be characterized using interchange entropies. ..."
-- [doi:10.1109/DCC50243.2021.00061]['21],
2011.04069@[arXiv]['21].
[Also search for: DCC21].
%A T. Gagie
%A G. Navarro
%A N. Prezza
%T Fully functional suffix trees and optimal text searching in BWT-runs bounded
space
%J JACM
%V 67
%N 1
%P 1-54
%M JAN
%D 2020
%K Jrnl, JACM, c020, c202x, c20xx, zz0321, run length FM index, FMindex, BWT,
RLBWT, suffix tree, repetitive strings, stringology
%X "Indexing highly repetitive text - such as genomic databases, s/w
repositories & versioned text collections ... One of the earliest indexes for
rep.collections, the Run-Length FM-index, used O(r) space & was able to
efficiently count the # of occurrences of a pattern of length m in a text of
length n (in O(m log log n) time, with current techniques). However, it was
unable to locate the posns of those occurrences efficiently within a space
bounded in terms of r. ... we close this long-standing problem, showing how
to extend the Run-Length FM-index so that it can locate the occ occurrences
efficiently (in O(occ log log n) time) within O(r) space. By raising the
space to O(r log log n), our index counts the occurrences in optimal time,
O(m), & locates them in optimal time as well, O(m+occ). By further raising
the space by an O(w/log s) factor, where s is the alphabet size &
w=Omega(log n) is the RAM machine size in bits, we support count & locate in
O(ceil(m log (s)/w)) & O(ceil(m log (s)/w)+occ) time, which is optimal in the
packed setting & had not been obtained before in compressed space. We also
describe a structure using O(r log (n/r)) space that replaces the text &
extracts any text substring of length L in the almost-optimal time
O(log (n/r)+L log (s)/w). Within that space, we similarly provide access to
arbitrary suffix array, inverse suffix array, & longest common prefix array
cells in time O(log(n/r)), and extend these capabilities to full suffix tree
functionality, typically in O(log(n/r)) time per operation. Our experiments
show that our O(r)-space index outperforms the space-competitive alternatives
by 1 to 2 orders of magnitude in time. Competitive implementations of the
original FM-index are outperformed by 1 to 2 orders of magnitude in space
&/or 2 to 3 in time.
-- [doi:10.1145/3375890]['21].
%A S. J. Puglisi
%A B. Zhukova
%T Smaller RLZ-compressed suffix arrays
%J Data Comp. Conf. (DCC)
%I IEEE
%M MAR
%D 2021
%K conf, c2021, c202x, c20xx, zz0321, DCC, DCC21, DCC2021, Snowbird, strings,
suffix array, index, RLZ, RLZSA, SAdiff
%X "Recently it was shown (auths SPIRE, 2020) that the suffix array (SA) data
structure can be effectively compressed with relative Lempel-Ziv (RLZ)
dictionary comp'n in such a way that arbitrary subarrays can be rapidly
decompressed, thus facilitating compressed indexing. ... we describe
optimizations to RLZ-compressed SAs, including generation of more effective
dictionaries & compact encodings of index components, both of which reduce
index size without adversely affecting subarray access speeds relative to
other compressed indexes. Our experimental analysis also elucidates the
relationship between subarray size and per elt access time.
-- [doi:0.1109/DCC50243.2021.00029]['21].
[Also search for: DCC21].