The MinMax k-Means clustering algorithm

Applying k-Means to minimize the sum of the intra-cluster variances is the most popular clustering approach. However, after a bad initialization, poor local optima can be easily obtained. To tackle the initialization problem of k-Means, we propose the MinMax k-Means algorithm, a method that assigns weights to the clusters relative to their variance and optimizes a weighted version of the k-Means objective. Weights are learned together with the cluster assignments, through an iterative procedure. The proposed weighting scheme limits the emergence of large variance clusters and allows high quality solutions to be systematically uncovered, irrespective of the initialization. Experiments verify the effectiveness of our approach and its robustness over bad initializations, as it compares favorably to both k-Means and other methods from the literature that consider the k-Means initialization problem.     

E-commerce transactions in a virtual environment: virtual transactions

E-commerce is a fundamental method of doing business, such that for a firm to say it is trading at all in the modern market-place it must have some element of on-line presence. Coupled with this is the explosion of the “population” of Massively Multiplayer On-line Role Playing Games and other shared virtual environments. Many suggest this will lead to a further dimension of commerce: virtual commerce. We discuss here the issues, current roadblocks and present state of an e-commerce transaction carried out completely within a virtual environment; a virtual transaction. Although technically such transactions are in a sense trivial, they raise many other issues in complex ways thus making V-transactions a highly interesting cross-disciplinary issue. We also discuss the social, ethical and regulatory implications for the virtual communities in these environments of such v-transactions, how their implementation affects the nature and management of a virtual environment, and how they represent a fundamental merging of the real and virtual worlds for the purpose of commerce. We highlight the minimal set of features a v-transaction capable virtual environment requires and suggest a model of how in the medium term they could be carried out via a methodology we call click-through, and that the developers of such environments will need to take on the multi-modal behavior of their users, as well as elements of the economic and political sciences in order to fully realize the commercial potential of the v-transaction.     

A Greedy EM Algorithm for Gaussian Mixture Learning

Learning a Gaussian mixture with a local algorithm like EM can be difficult because (i) the true number of mixing components is usually unknown, (ii) there is no generally accepted method for parameter initialization, and (iii) the algorithm can get trapped in one of the many local maxima of the likelihood function. In this paper we propose a greedy algorithm for learning a Gaussian mixture which tries to overcome these limitations. In particular, starting with a single component and adding components sequentially until a maximum number k, the algorithm is capable of achieving solutions superior to EM with k components in terms of the likelihood of a test set. The algorithm is based on recent theoretical results on incremental mixture density estimation, and uses a combination of global and local search each time a new component is added to the mixture. This revised version was published online in August 2006 with corrections to the Cover Date.     

Copolymers of anionic polymerization of octanelactam with laurolactam. II. Kinetics of isothermal crystallization

The isothermal crystallization kinetics of the anionic copolyamides of ω-octanelactam (OL) with ω-laurolactam (LL) (nylon-8/nylon-12), covering a wide composition range (0/100 – 100/0 OL/LL, mol/mol), was established according to the Avrami equation by the use of differential scanning calorimetry. The crystallization parameters half-time crystallization (t_1/2), Avrami index (n), rate constant (k), and reduced crystallinity were determined for various crystallization temperatures. These parameter values are correlated to the composition of the copolyamides, the crystallization and supercooling temperature, and the results are discussed. Differences in kinetic parameter values between the homopolymers of this paper and those for the hydrolytic nylon-8 and nylon-12 reported in literature were observed and discussed.     

Variational Bayesian Image Restoration Based on a Product of t-Distributions Image Prior

Image priors based on products have been recognized to offer many advantages because they allow simultaneous enforcement of multiple constraints. However, they are inconvenient for Bayesian inference because it is hard to find their normalization constant in closed form. In this paper, a new Bayesian algorithm is proposed for the image restoration problem that bypasses this difficulty. An image prior is defined by imposing Student-t densities on the outputs of local convolutional filters. A variational methodology, with a constrained expectation step, is used to infer the restored image. Numerical experiments are shown that compare this methodology to previous ones and demonstrate its advantages.     

Training Reinforcement Neurocontrollers Using the Polytope Algorithm

A new training algorithm is presented for delayed reinforcement learning problems that does not assume the existence of a critic model and employs the polytope optimization algorithm to adjust the weights of the action network so that a simple direct measure of the training performance is maximized. Experimental results from the application of the method to the pole balancing problem indicate improved training performance compared with critic-based and genetic reinforcement approaches.     

Kriechbeiwerte für die Berechnung von Verbundträgern

Creep factors for the computation of composite girders. Cross‐section properties, deformations and stresses in composite girders under consideration of creep and shrinkage are mainly calculated with the help of modular ratios. The application of this method demands the knowledge of creep factors, which take into account the long term behaviour of concrete and the time depended effects of the stresses in concrete. In literature there are different values for creep factors. In this article the precision of these factors and the influence of deviations are investigated. Proposals for the calculation procedure of creep and shrinkage effects are summarized.     

Anionic copolymers of octanelactam with laurolactam (nylon-8/12)—IV. Dynamic thermomechanical and mechanical properties

The thermomechanical and the mechanical properties of a recent series of copolyamides of octanelactam (OL) and laurolactam (LL) (nylon-8/nylon-12) were studied. The glass transition temperature (T_g) and melting point (T_m) were determined with dynamic mechanical thermal analyser, and compared to those obtained with differential scanning calorimetry and the activation energy for glass transition was calculated. The copolyamides showed a minimum for tensile strength, yield stress, stress at 100% and modulus and a maximum for elongation at break at the composition 60/40 (OL/LL) which has the lowest crystallinity. Young's modulus against % elongation showed the classification of copolyamides in two groups (rich in OL or LL, respectively).     

Preparation and characterization of lyophilised egg PC liposomes incorporating curcumin and evaluation of its activity against colorectal cancer cell lines

Curcumin has been associated with the treatment of various diseases in traditional medicine, among them cancer. The major problems that prevent its approval as therapeutic agent are its low water solubility and its relatively low in vivo bioavailability. Liposomes are considered as effective drug carriers because of their ability to solubilize hydrophobic compounds and to alter their pharmacokinetic properties. The purpose of this study was the development of lyophilised liposomal curcumin fully characterized in terms of its physical properties [(zeta-potential, size, size distribution and Polydispercity index (PI)], and to evaluate its in vitro cytotoxic against colorectal cancer cell lines in a short-term and in a long-term (clonogenic) assay. Curcumin was incorporated in egg-phosphatidylcholine (EPC) liposomes at a drug to lipid molar ratio 1:14 achieving high incorporation efficiency close to 85%. The liposomal curcumin was lyophilized preserving thus its stability. The reconstitution of the formulation resulted in the original liposomal suspension. The release in FBS showed a plateau near 14% at 96 hours of incubation. The in vitro studies against colorectal cancer cell lines have shown that liposomes improve the activity of curcumin especially in the long-term assay and the liposomal formulation found to be more potent against HCT116 and HCT15, cell lines which express the MDR phenotype. EPC liposomal curcumin in a molar ratio of curcumin/EPC 1:14 has shown improved cytotoxic activity versus free curcumin against colorectal cancer cell lines. In vivo studies based on the recent findings are in progress in our laboratory.