Turbo-smt: Accelerating coupled sparse matrix-tensor factorizations by 200x

How can we correlate the neural activity in the human brain as it responds to typed words, with properties of these terms (like ‘edible’, ‘fits in hand’)? In short, we want to find latent variables, that jointly explain both the brain activity, as well as the behavioral responses. This is one of many settings of the Coupled Matrix-Tensor Factorization (CMTF) problem.

Can we accelerate any CMTF solver, so that it runs within a few minutes instead of tens of hours to a day, while maintaining good accuracy? We introduce Turbo-SMT, a meta-method capable of doing exactly that: it boosts the performance of any CMTF algorithm, by up to 200x, along with an up to 65 fold increase in sparsity, with comparable accuracy to the baseline.

We apply Turbo-SMT to BrainQ, a dataset consisting of a (nouns, brain voxels, human subjects) tensor and a (nouns, properties) matrix, with coupling along the nouns dimension. Turbo-SMT is able to find meaningful latent variables, as well as to predict brain activity with competitive accuracy.

Extended factorizations of exponential functionals of Lévy processes

Pardo, Patie, and Savov derived, under mild conditions, a Wiener-Hopf type factorization for the exponential functional of proper Lévy processes. In this paper, we extend this factorization by relaxing a finite moment assumption as well as by considering the exponential functional for killed Lévy processes. As a by-product, we derive some interesting fine distributional properties enjoyed by a large class of this random variable, such as the absolute continuity of its distribution and the smoothness, boundedness or complete monotonicity of its density. This type of results is then used to derive similar properties for the law of maxima and first passage time of some stable Lévy processes. Thus, for example, we show that for any stable process with $\rho\in(0,\frac{1}{\alpha}-1]$, where $\rho\in[0,1]$ is the positivity parameter and $\alpha$ is the stable index, then the first passage time has a bounded and non-increasing density on $\mathbb{R}_+$. We also generate many instances of integral or power series representations for the law of the exponential functional of Lévy processes with one or two-sided jumps. The proof of our main results requires different devices from the one developed by Pardo, Patie, Savov. It relies in particular on a generalization of a transform recently introduced by Chazal et al together with some extensions to killed Lévy process of Wiener-Hopf techniques. The factorizations developed here also allow for further applications which we only indicate here also allow for further applications which we only indicate here.

Bayesian versus frequentist statistical inference for investi gating a one-off cancer cluster reported to a health department

Background The problem of silent multiple comparisons is one of the most difficult statistical problems faced by scientists. It is a particular problem for investigating a one-off cancer cluster reported to a health department because any one of hundreds, or possibly thousands, of neighbourhoods, schools, or workplaces could have reported a cluster, which could have been for any one of several types of cancer or any one of several time periods. Methods This paper contrasts the frequentist approach with a Bayesian approach for dealing with silent multiple comparisons in the context of a one-off cluster reported to a health department. Two published cluster investigations were re-analysed using the Dunn-Sidak method to adjust frequentist p-values and confidence intervals for silent multiple comparisons. Bayesian methods were based on the Gamma distribution. Results Bayesian analysis with non-informative priors produced results similar to the frequentist analysis, and suggested that both clusters represented a statistical excess. In the frequentist framework, the statistical significance of both clusters was extremely sensitive to the number of silent multiple comparisons, which can only ever be a subjective "guesstimate". The Bayesian approach is also subjective: whether there is an apparent statistical excess depends on the specified prior. Conclusion In cluster investigations, the frequentist approach is just as subjective as the Bayesian approach, but the Bayesian approach is less ambitious in that it treats the analysis as a synthesis of data and personal judgements (possibly poor ones), rather than objective reality. Bayesian analysis is (arguably) a useful tool to support complicated decision-making, because it makes the uncertainty associated with silent multiple comparisons explicit.

Sodium 2-nitrocinnamate dihydrate: a one-dimensional hydrogen-bonded coordination polymer

The title compound catena-poly[aqua-mu3-2-nitrocinnamato], [Na(C9H6NO4)(H2O)2]n, the sodium salt of trans-2-nitrocinnamic acid, is a one-dimensional coordination polymer based on six-coordinate octahedral NaO6 centres comprising three facially-related monodentate carboxylate O-atom donors from separate ligands (all bridging)[Na-O, 2.4370(13)-2.5046(13)A] and three water molecules (two bridging, one monodentate) [Na-O, 2.3782(13)-2.4404(17)A]. The structure is also stabilized by intra-chain water-O-H...O(carboxylate) and O-H...O(nitro) hydrogen bonds.

Zero- one- and two-dimensional hydrogen-bonded structures in the 1:1 proton-transfer compounds of 4,5-dichlorophthalic acid with the monocyclic heteroaromatic Lewis bases 2-aminopyrimidine, nicotinamide and isonicotinamide

The structures of the anhydrous 1:1 proton-transfer compounds of 4,5-dichlorophthalic acid (DCPA) with the monocyclic heteroaromatic Lewis bases 2-aminopyrimidine, 3-(aminocarboxy) pyridine (nicotinamide) and 4-(aminocarbonyl) pyridine (isonicotinamide), namely 2-aminopyrimidinium 2-carboxy-4,5-dichlorobenzoate C4H6N3+ C8H3Cl2O4- (I), 3-(aminocarbonyl) pyridinium 2-carboxy-4,5-dichlorobenzoate C6H7N2O+ C8H3Cl2O4- (II) and the unusual salt adduct 4-(aminocarbonyl) pyridinium 2-carboxy-4,5-dichlorobenzoate 2-carboxymethyl-4,5-dichlorobenzoic acid (1/1/1) C6H7N2O+ C8H3Cl2O4-.C9H6Cl2O4 (III) have been determined at 130 K. Compound (I) forms discrete centrosymmetric hydrogen-bonded cyclic bis(cation--anion) units having both R2/2(8) and R2/1(4) N-H...O interactions. In compound (II) the primary N-H...O linked cation--anion units are extended into a two-dimensional sheet structure via amide-carboxyl and amide-carbonyl N-H...O interactions. The structure of (III) reveals the presence of an unusual and unexpected self-synthesized methyl monoester of the acid as an adduct molecule giving one-dimensional hydrogen-bonded chains. In all three structures the hydrogen phthalate anions are

Knowing a donor and identifying as one : determinants of people's willingness for related and anonymous living donation in Australia

While LRD (living donation to a genetically/emotionally related recipient) is well established in Australia, LAD (living anonymous donation to a stranger) is rare. Given the increasing use of LAD overseas, Australia may likely follow suit. Understanding the determinants of people’s willingness for LAD is essential but infrequently studied in Australia. Consequently, we compared the determinants of people’s LRD and LAD willingness, and assessed whether these determinants differed according to type of living donation. We surveyed 487 health students about their LRD and LAD willingness, attitudes, identity, prior experience with blood and organ donation, deceased donation preference, and demographics. We used Structural Equation Modelling (SEM) to identify the determinants of willingness for LRD and LAD and paired sample t-tests to examine differences in LRD and LAD attitudes, identity, and willingness. Mean differences in willingness (LRD 5.93, LAD 3.92), attitudes (LRD 6.43, LAD 5.53), and identity (LRD 5.69, LAD 3.58) were statistically significant. Revised SEM models provided a good fit to the data (LRD: x2 (41) = 67.67, p = 0.005, CFI = 0.96, RMSEA = 0.04; LAD: x2 (40) = 79.64, p < 0.001, CFI = 0.95, RMSEA = 0.05) and explained 45% and 54% of the variation in LRD and LAD willingness, respectively. Four common determinants of LRD and LAD willingness emerged: identity, attitude, past blood donation, and knowing a deceased donor. Religious affiliation and deceased donation preference predicted LAD willingness also. Identifying similarities and differences in these determinants can inform future efforts aimed at understanding people’s LRD and LAD willingness and the evaluation of potential living donor motives. Notably, this study highlights the importance of people’s identification as a living donor as a motive underlying their willingness to donate their organs while living.

One size doesn't fit all. Exploring marketing strategies for influencing three dimensions of brand loyalty

Brand loyalty is a concept that has garnered considerable interest over recent years from both marketing practitioners and academics alike. While marketers are primarily interested in ways they can generate and increase brand loyalty from their customers, academics strive to conducts research which investigates the antecedents and consequences of customer loyalty (See DeWitt, Nguyen and Marshall 2008; Russell-Bennett, McColl-Kennedy and Coote 2007).

Efficient one-round key exchange in the standard model

We consider one-round key exchange protocols secure in the standard model. The security analysis uses the powerful security model of Canetti and Krawczyk and a natural extension of it to the ID-based setting. It is shown how KEMs can be used in a generic way to obtain two different protocol designs with progressively stronger security guarantees. A detailed analysis of the performance of the protocols is included; surprisingly, when instantiated with specific KEM constructions, the resulting protocols are competitive with the best previous schemes that have proofs only in the random oracle model.

Efficient one-round key exchange in the efficient one-round key exchange in the standard model

We consider one-round key exchange protocols secure in the standard model. The security analysis uses the powerful security model of Canetti and Krawczyk and a natural extension of it to the ID-based setting. It is shown how KEMs can be used in a generic way to obtain two different protocol designs with progressively stronger security guarantees. A detailed analysis of the performance of the protocols is included; surprisingly, when instantiated with specific KEM constructions, the resulting protocols are competitive with the best previous schemes that have proofs only in the random oracle model.