Refinements of universal approximation results for deep belief networks and restricted Boltzmann machines

We improve recently published results about resources of restricted Boltzmann machines (RBM) and deep belief networks (DBN)required to make them universal approximators. We show that any distribution pon the set {0,1}(n) of binary vectors of length n can be arbitrarily well approximated by an RBM with k-1 hidden units, where k is the minimal number of pairs of binary vectors differing in only one entry such that their union contains the support set of p. In important cases this number is half the cardinality of the support set of p (given in Le Roux & Bengio, 2008). We construct a DBN with 2n/ 2(n-b) , b ～ log n, hidden layers of width n that is capable of approximating any distribution on {0,1}(n) arbitrarily well. This confirms a conjecture presented in Le Roux and Bengio (2010).     

Recurrent type-1 fuzzy functions approach for time series forecasting

Forecasting the future values of a time series is a common research topic and is studied using probabilistic and non-probabilistic methods. For probabilistic methods, the autoregressive integrated moving average and exponential smoothing methods are commonly used, whereas for non-probabilistic methods, artificial neural networks and fuzzy inference systems (FIS) are commonly used. There are numerous FIS methods. While most of these methods are rule-based, there are a few methods that do not require rules, such as the type-1 fuzzy function (T1FF) approach. While it is possible to encounter a method such as an autoregressive (AR) model integrated with a T1FF, no method that includes T1FF and the moving average (MA) model in one algorithm has yet been proposed. The aim of this study is to improve forecasting by taking the disturbance terms into account. The input dataset is organized using the following variables. First, the lagged values of the time series are used for the AR model. Second, a fuzzy c-means clustering algorithm is used to cluster the inputs. Third, for the MA, the residuals of fuzzy functions are used. Hence, AR, MA, and the degree of memberships of the objects are included in the input dataset. Because the objective function is not derivative, particle swarm optimization is preferable for solving it. The results on several datasets show that the proposed method outperforms most of the methods in literature.     

Distribution of uranium in the production of triple superphosphate (TSP) fertilizer and phosphoric acid

In this study the distribution of uranium, which is one of the radioactive elements present in phosphate fertilizers was investigated in different steps of the triple superphosphate (TSP) production process. The uranium in phosphate rock, phosphoric acid, phosphogypsum and TSP was extracted into the organic phase by using the TOPO extraction method. The uranium contents of these materials were determined by measuring the absorbance of the formed (pH 7.6) uranyl bromo-PADAP [uranyl(2-(5-bromo-2-pyridylazo)-5-diethylaminophenol)] complex solutions at 574 nm against blank. It was found that 50% of the uranium is dissolved in the acid during the production of phosphoric acid while the remainder is precipitated with phosphogypsum residue. The observations showed that in the second step, the sum of uranium in phosphate rock and phosphoric acid completely passed into TSP in the TSP manufacturing process.     

Analyzing the effect of the elastane-containing dual-core weft yarn density on the denim fabric performance properties

Denim fabrics form today’s mostly utilized fabric type. As is the case with the other textile products, there are many factors affecting the properties and the performance of the denim products. Within the scope of this study, we have evaluated the effect of the density changes in the use of the dual-core threads – used in an ever-increasing fashion in the textile industry – in weft have on the fabric properties. We have analyzed the extent to which the weight, size, elasticity, tensile strength, and cost properties of the denim fabrics woven with the dual-core weft thread in various densities are affected by the changes in the number of dual-core weft threads per unit length. In conclusion, we have come to such striking remarks as that the construction has a much more impact on the fabric width and thus on the unit weight than the elasticity ratio, and that density changes in the elastane-containing threads cause serious differences on the fabric’s color values.     

Effect of prewetting methods on some fresh and hardened properties of concrete with pumice aggregate

One of the major problems in lightweight aggregate concrete production is the high water absorption characteristic of the aggregates due to their porous structure. This problem is usually overcome by prewetting the lightweight aggregates or increasing the amount of mixing water. Since aggregate prewetting methods significantly affect fresh and hardened lightweight concrete properties, it is important to take this into account before the concrete production process.This study is focused on the effects of three prewetting methods on some fresh and hardened properties of pumice lightweight concrete. Pre-soaking, water-soaking and vacuum-soaking methods were applied to pumice lightweight aggregate prior to mixing. Test results showed that fresh and hardened properties of concretes with vacuum-soaked and water-soaked lightweight aggregate were significantly better than that of concretes with pre-soaked lightweight aggregate. Vacuum-soaking and water-soaking of pumice aggregate improved workability, compressive strength and drying shrinkage of pumice lightweight concrete.     

A Novel ARMA Type Possibilistic Fuzzy Forecasting Functions Based on Grey-Wolf Optimizer (ARMA-PFFs)

Computational Economics - This study proposes a new time series forecasting method that employs possibilistic fuzzy c-means, an autoregressive moving average model (ARMA), and a grey wolf optimizer...     

Discharge Coefficient for Trapezoidal Labyrinth Side Weir in Subcritical Flow

The discharge coefficient of a trapezoidal labyrinth side weir is a function of the Froude number F ₁, the dimensionless effective crest length L/ℓ, the dimensionless weir length L/B, the dimensionless weir height p/h ₁, and the sidewall angle α. A labyrinth weir is an overflow weir, folded in plan view to provide a longer total effective length for a given overall weir width. These weirs have advantages compared to the straight overflow weir and the standard ogee crest. Previous studies on the subject have generally focused on rectangular side weirs located on a straight channel. The present study investigates the hydraulic behavior of a trapezoidal labyrinth side weir. The results show that the discharge coefficient of labyrinth side weirs gives a significantly higher coefficient value compare to that of conventional straight side weirs. Discharge coefficient of the trapezoidal labyrinth side weir is 1.5 to 5.0 times higher than the conventional straight side weir. Consequently, an equation for the coefficient of discharge is introduced. The results predicted by the equation were shown to be very satisfactory using root mean square error (RMSE), mean absolute error (MAE) and correlation coefficient (R) statistics.     

Modeling multilayered MEMS-based micro-fluidic systems

This article describes the design, fabrication, and simulation of two micro-electromechanical systems-based micro-fluidic systems. The first system, a lab-on-a-chip, enables electrochemical immunoassay-based chemical/biological detection. The second is a micro-fluidic bio-impedance sensor. The relevance of fluidic dynamics in micro-fluidic channels is discussed in context of fluid paths and misalignments in the channels that appear during multi-level structure integration. Also discussed is the effect of channel dimensions on the flow profile and on performance within the micro-systems. For more information, contact Shekhar Bhansali, University of South Florida, Nanomaterials and Nanomanufacturing Research Center, Department of Electrical Engineering, 4202 East Fowler Avenue ENB 118, Tampa, Florida 33620; e-mail bhansali@eng.usf.edu.