Analysis of financial stock markets through multidimensional scaling based on information measures

In this paper, a novel multidimensional scaling (MDS) based on information measures method is proposed to analyze financial stock markets. In order to examine the effectiveness of this method, we applied it to the classification of two types of artificial series, the logistic map model and the cubic map model, as well as stock time series. Moreover, the traditional MDS using Euclidean dissimilarity is also provided as a reference for comparisons. The results show that the MDS based on information measures can give us more detailed, exact and clearer information on the classification of simulation series and stock time series than the MDS using Euclidean dissimilarity. In addition, the proposed graphical method may also assist in the construction of multivariate econometric models.     

Unsupervised learning of pharmacokinetic responses

Pharmacokinetics (PK) is a branch of pharmacology dedicated to the study of the time course of drug concentrations, from absorption to excretion from the body. PK dynamic models are often based on homogeneous, multi-compartment assumptions, which allow to identify the PK parameters and further predict the time evolution of drug concentration for a given subject. One key characteristic of these time series is their high variability among patients, which may hamper their correct stratification. In the present work, we address this variability by estimating the PK parameters and simultaneously clustering the corresponding subjects using the time series. We propose an expectation maximization algorithm that clusters subjects based on their PK drug responses, in an unsupervised way, collapsing clusters that are closer than a given threshold. Experimental results show that the proposed algorithm converges fast and leads to meaningful results in synthetic and real scenarios.     

Targeting Bayes factors with direct-path non-equilibrium thermodynamic integration

Thermodynamic integration (TI) for computing marginal likelihoods is based on an inverse annealing path from the prior to the posterior distribution. In many cases, the resulting estimator suffers from high variability, which particularly stems from the prior regime. When comparing complex models with differences in a comparatively small number of parameters, intrinsic errors from sampling fluctuations may outweigh the differences in the log marginal likelihood estimates. In the present article, we propose a TI scheme that directly targets the log Bayes factor. The method is based on a modified annealing path between the posterior distributions of the two models compared, which systematically avoids the high variance prior regime. We combine this scheme with the concept of non-equilibrium TI to minimise discretisation errors from numerical integration. Results obtained on Bayesian regression models applied to standard benchmark data, and a complex hierarchical model applied to biopathway inference, demonstrate a significant reduction in estimator variance over state-of-the-art TI methods.     

A Cut-Invariant Law of Large Numbers for Random Heaps

We consider the framework of Bernoulli measures for heap monoids. We introduce in this framework the notion of asynchronous stopping time, which generalizes the notion of stopping time for classical probabilistic processes. A strong Bernoulli property is proved. A notion of cut-invariance is formulated for convergent ergodic means. Then, a version of the strong law of large numbers is proved for heap monoids with Bernoulli measures. We study a sub-additive version of the law of large numbers in this framework.     

Fuzzy identification of cutting acoustic emission with extended subtractive cluster analysis

This paper presents fuzzy acoustic emission identification in high precision hard turning process based on extended subtractive cluster analysis combined with the least-square estimation method. The fuzzy identification method provides a simple way to arrive at a definite conclusion based upon the information obtained with the difficulty in understanding the exact physics of the machining process. The experimental results prove that the proposed method is efficient and feasible.     

Ultrathin functional films of titanium(IV) oxide

Chemistry and physics of thin semiconducting layers of various types are subjects of intense research. Especially when nanotechnology methods such as self-assembly are involved, amazing structural and/or functional properties may appear. Also modern physical methods using variously organized plasma arrangements are able to produce uniform structures with distinctive functionality. In this review, based virtually on our own work, discussions on the preparation, structure, morphology, and function of titanium(IV) oxide nanoscopic thin films are presented. It was shown that structurally and functionally similar titanium(IV) oxide films can be prepared via completely different preparation techniques. Function tests were arranged as “primary”, covering the assessment of the light induced charge separation efficiency, and “secondary”, based on photocatalytic surface oxidations.