Learning Effective and Interpretable Semantic Models using Non-Negative Sparse Embedding

In this paper, we introduce an application of matrix factorization to produce corpus-derived, distributional
models of semantics that demonstrate cognitive plausibility. We find that word representations
learned by Non-Negative Sparse Embedding (NNSE), a variant of matrix factorization, are sparse,
effective, and highly interpretable. To the best of our knowledge, this is the first approach which
yields semantic representation of words satisfying these three desirable properties. Though extensive
experimental evaluations on multiple real-world tasks and datasets, we demonstrate the superiority
of semantic models learned by NNSE over other state-of-the-art baselines.

Discovering subgroups of patients from DNA copy number data using NMF on compacted matrices

In the study of complex genetic diseases, the identification of subgroups of patients sharing similar genetic characteristics represents a challenging task, for example, to improve treatment decision. One type of genetic lesion, frequently investigated in such disorders, is the change of the DNA copy number (CN) at specific genomic traits. Non-negative Matrix Factorization (NMF) is a standard technique to reduce the dimensionality of a data set and to cluster data samples, while keeping its most relevant information in meaningful components. Thus, it can be used to discover subgroups of patients from CN profiles. It is however computationally impractical for very high dimensional data, such as CN microarray data. Deciding the most suitable number of subgroups is also a challenging problem. The aim of this work is to derive a procedure to compact high dimensional data, in order to improve NMF applicability without compromising the quality of the clustering. This is particularly important for analyzing high-resolution microarray data. Many commonly used quality measures, as well as our own measures, are employed to decide the number of subgroups and to assess the quality of the results. Our measures are based on the idea of identifying robust subgroups, inspired by biologically/clinically relevance instead of simply aiming at well-separated clusters. We evaluate our procedure using four real independent data sets. In these data sets, our method was able to find accurate subgroups with individual molecular and clinical features and outperformed the standard NMF in terms of accuracy in the factorization fitness function. Hence, it can be useful for the discovery of subgroups of patients with similar CN profiles in the study of heterogeneous diseases.

Scoup-SMT: Scalable Coupled Sparse Matrix-Tensor Factorization

How can we correlate neural activity in the human brain as it responds to words, with behavioral data expressed as answers to questions about these same words? In short, we want to find latent variables, that explain both the brain activity, as well as the behavioral responses. We show that this is an instance of the Coupled Matrix-Tensor Factorization (CMTF) problem. We propose Scoup-SMT, a novel, fast, and parallel algorithm that solves the CMTF problem and produces a sparse latent low-rank subspace of the data. In our experiments, we find that Scoup-SMT is 50-100 times faster than a state-of-the-art algorithm for CMTF, along with a 5 fold increase in sparsity. Moreover, we extend Scoup-SMT to handle missing data without degradation of performance. We apply Scoup-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. Scoup-SMT is able to find meaningful latent variables, as well as to predict brain activity with competitive accuracy. Finally, we demonstrate the generality of Scoup-SMT, by applying it on a Facebook dataset (users, friends, wall-postings); there, Scoup-SMT spots spammer-like anomalies.

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.

In vitro release of nonoxynol-9 from silicone matrix intravaginal rings

The controlled-release characteristics of matrix silicone intravaginal rings loaded with between 100 and 971 mg of nonoxynol-9 have been investigated with a view to developing a ring that may offer a new female-controlled method for the prevention of transmission of sexually transmitted diseases, particularly HIV. Intravaginal rings containing 253, 487 and 971 mg of nonoxynol-9 provided a daily release of 2 mg or more over the 8-day release period, the minimal amount of nonoxynol-9 considered to provide an effective vaginal concentration for the prevention of HIV. Furthermore, the maximum daily release of N9 was about 6 mg, an amount significantly smaller than that observed for other nonoxynol-9 products whose large daily doses may in fact increase the occurrence of HIV by causing epithelial damage to the vaginal tissue. The release mechanism of the liquid nonoxynol-9 from the intravaginal rings has also been investigated and compared to models describing the release of solid drugs from the rings. It has been demonstrated through release studies and surface microscopy that a drug depletion zone is not established in such liquid-loaded intravaginal ring systems, with implications for the release kinetics. (C) 2003 Elsevier B.V. All rights reserved.

High speed DSC (hyper-DSC) as a tool to measure the solubility of a drug within a solid or semi-solid matrix

Conventional differential scanning calorimetry (DSC) techniques are commonly used to quantify the solubility of drugs within polymeric-controlled delivery systems. However, the nature of the DSC experiment, and in particular the relatively slow heating rates employed, limit its use to the measurement of drug solubility at the drug's melting temperature. Here, we describe the application of hyper-DSC (HDSC), a variant of DSC involving extremely rapid heating rates, to the calculation of the solubility of a model drug, metronidazole, in silicone elastomer, and demonstrate that the faster heating rates permit the solubility to be calculated under non-equilibrium conditions such that the solubility better approximates that at the temperature of use. At a heating rate of 400 degrees C/min (HDSC), metronidazole solubility was calculated to be 2.16 mg/g compared with 6.16 mg/g at 20 degrees C/min. (C) 2005 Elsevier B.V. All rights reserved.

Children with stable asthma have reduced airway matrix metalloproteinase-9 and matrix metalloproteinase-9/tissue inhibitor of metalloproteinase-1 ratio

Background Childhood asthma is characterized by inflammation of the airways. Structural changes of the airway wall may also be seen in some children early in the course of the disease. Matrix metalloproteinases (MMPs) are key mediators in the metabolism of the extracellular matrix (ECM). Objective To investigate the balance of MMP-8, MMP-9 and tissue inhibitor of metalloproteinases (TIMP)-1 in the airways of children with asthma. Methods One hundred and twenty-four children undergoing elective surgical procedures also underwent non-bronchoscopic bronchoalveolar lavage (BAL). MMP-8, MMP-9 and TIMP-1 were measured by ELISA. Results There was a significant reduction in MMP-9 in atopic asthmatic children (n=31) compared with normal children (n=30) [median difference: 0.57 ng/mL (95% confidence interval: 0.18–1.1 ng/mL)]. The ratio of MMP-9 to TIMP-1 was also reduced in asthmatic children. Levels of all three proteins were significantly correlated to each other and to the relative proportions of particular inflammatory cells in BAL fluid (BALF). Both MMP-8 and MMP-9 were moderately strongly correlated to the percentage neutrophil count (r=0.40 and 0.47, respectively, P