Soil temperature has a vital importance in biological, physical and chemical processes of terrestrial ecosystem its modeling at different depths is very important for land-atmosphere interactions. The study compares four machine learning techniques, extreme (ELM), artificial neural networks (ANN), classification regression trees (CART) group method data handling (GMDH) estimating monthly soil temperatures depths. Various combinations climatic variables are utilized as input to the developed...

Abstract Accurately modeling pan evaporation is important in water resources planning and management also environmental engineering. This study compares the accuracy of two new data-driven methods, multi-gene genetic programming (MGGP) approach dynamic evolving neural-fuzzy inference system (DENFIS), monthly evaporation. The climatic data, namely, minimum temperature, maximum solar radiation, relative humidity, wind speed, evaporation, obtained from Antakya Antalya stations, Mediterranean...

Aim of this study is applying the ensemble classification methods over stock market closing values, which can be assumed as time series and finding out relation between economy news. In order to keep back ground clear, majority voting method has been applied three algorithms, are k-nearest neighborhood, support vector machine C4.5 tree. The results gathered from two different feature extraction correlated with meta classifier (ensemble method) running classifiers. show success rates...

This paper proposes an information retrieval method for the economy news. The effect of news, are researched in word level and stock market values considered as ground proof. correlation between prices news is already addressed problem most countries. well-known approach applying text mining approaches to some time series analysis techniques over closing order apply classification or clustering algorithms features extracted. study goes further tries ask question what available which one...

This study aims to publish a novel similarity metric increase the speed of comparison operations.Also new is suitable for distance-based operations among strings.Most simple calculation methods, such as string length are fast calculate but doesn't represent correctly.On other hand methods like keeping histogram over all characters in slower good characteristics some areas, natural language.We propose metric, easy and satisfactory comparison.Method built on hash function, which gets at any...

An application of Singular Spectrum Analysis(SSA) Method, based on a new elaborated tensorial approach computation singular values and left right vectors arbitrary non-square matrices, for time series is presented. All necessary calculations both types (left right) are performed the base approach. It showed that non parametric SSA can be efficiently used as universal filter to separate Low High frequencies components in long signals series.

This study focuses on the second group of hashing algorithms and criticizes using Feistel Networkwhich are widely utilized in text mining studies. We propose a new approach which is mainly built substitution boxes (sboxes),which core all Networks processes faster than other implementations.

Depending on the market strength and structure, it is a known fact that there correlation between stock values content in newspapers. The increases weak speculative markets, while they never get reduced to zero strongest markets. This research focuses economic news published highly circulating newspaper Turkey closing Turkey. In several feature extraction methodologies are implemented both of data sources, which news. Since natural language format, text mining technique, term frequency –...

The study examines the trend analysis of temperatures in Tbilisi City, Georgiaby using non-parametric Mann Kendall test. Data covered monthly air temperatureof for period 1901-2012. Although no significant was seen whole temperature data according to 95% confidence level, some months showed significantly increasing trends. January, April, July and September while identified other with respect level. Annual were also analyzed a highly City.

ÖZETBu çalışmanın amacı, öncelikli olarak RSA şifreleme yönteminde kullanılan ve iki asal sayının çarpımından oluşan yarı-

ABSTRACT This research conducts the web statistics of employment sites with technological impact on macroeconomics. The statistical information gathered from web-o-metrics Caucasus region job seeking like number visitors, Facebook likes or shares, twitter messages about site, back links counted by google, bing Alexa. On other hand, macroeconomic and demographic facts population, unemployment rate, median age migration rate. Keywords Unemployment, Job Market, Cross-Country Data Mining, Time...

In this paper, we describe an algorithm for solving a certain system of algebraic linear equations using the underlying displacement structureof coefficients matrix system. Fast solution is key component acceleration recently developed novel matrixspectral factorization algorithm. The results numerical simulations, which compare optimized software implementation structured system’ssolution to standard one built in MATLAB, are presented as well.

Approximation and filtration of time series belongs to one most important problems in real word scientific application. Despite plenty ofexisted methods, they are effective very specific situations, them assume that is stationary or can be by finiteamount differencing. In this article we will consider several compare Singular spectrum analysis its low rank tensorialapproximation method classical wavelet decomposition approach.

Recently, in data science one of the most important issues has been discovering actionable information, interpretable patterns and relationshipsin large volumes data. This process is called mining commonly being used science, engineering, business security.One main methods similarity search time series. The approach that discussed this article based on PiecewiseLinear Representation series imply two steps measuring similarity. A new method piecewise linear approximationof non-stationary developed.

This study aims to publish a novel similarity metric increase the speed of comparison operations. Also new is suitable for distance-based operations among strings. Most simple calculation methods, such as string length are fast calculate but does not represent correctly. On other hand methods like keeping histogram over all characters in slower good characteristics some areas, natural language. We propose metric, easy and satisfactory comparison. Method built on hash function, which gets at...

The study investigates the ability of M5 model tree in modeling monthly solarradiation. Monthly data maximum and minimum air temperatures, windspeed, relative humidity solar radiation from Antalya, Turkey were used inthe application. effect each climatic parameter on is investigated. Periodic models are also developed to see periodicitycomponent models’ accuracy. It was found that using periodicity as input tothe applied significantly increases their accuracy monthlysolar radiation.