Browsing by Author "Soyupak, Selcuk"

Now showing 1 - 5 of 5

Citation - Scopus: 1
Adaptive Neuro-Fuzzy Inference Technique for Estimation of Light Penetration in Reservoirs
(Springer Japan Kk, 2007) Soyupak, Selcuk; Karaer, Feza; Senturk, Engin; Hekim, Huseyin; 01. Atılım University
An adaptive neuro-fuzzy inference technique has been adopted to estimate light levels in a reservoir. The data were collected randomly from Doganci Dam Reservoir over a number of years. The input data set is a matrix with vectors of time, depth, sampling location, and incident solar radiation. The output data set is a vector representing light measured at various depths. Randomization and logarithmic transformations have been applied as preprocessing. One-half of the data have been utilized for training; testing and validation steps utilized one-fourth each. An adaptive neuro-fuzzy inference system (ANFIS) has been built as a prediction model for light penetration. Very high correlation values between predictions and real values on light measurements with relatively low root mean square error values have been obtained for training, test, and validation data sets. Elimination of the overtraining problem was ensured by satisfying close root mean square error values for all sets.
Citation - WoS: 1
Application of Artificial Neural Network-Based Approach for Calculating Dissolved Oxygen Profiles in Kapulukaya Dam Reservoir
(Centre Environment Social & Economic Research Publ-ceser, 2007) Tuzun, Ilhami; Soyupak, Selcuk; Ince, Ozlem; Basaran, Gokben; 01. Atılım University
An Artificial Neural Network (ANN) modelling approach has been shown to be successful in calculating time and space dependent dissolved oxygen (DO) concentration profiles in Kapulukaya Dam Reservoir using limited number of input variables. The variation of inflow to the reservoir with respect to time was significantly high. The reservoir operational levels were relatively stable. The Levenberg-Marquardt algorithm was adopted during training. Preprocessing before training and post processing after simulation steps were the treatments applied to raw data and predictions respectively. Different configurations of Multilayer perceptron neural networks were designed by selecting different combinations of number of hidden layers (single and double) and number of neurons within each of the hidden layers. Generalisation was improved and over-fitting problems were eliminated: Early stopping method was applied for improving generalisation. The conventional model criteria of correlation coefficient (R) and mean square errors (MSE) were adopted to compare model performances. The correlation coefficients between neural network estimates and field measurements were as high as 0.96 for daily and monthly data respectively with experiments that involve double layer neural network structure with 31 neurons within each hidden layer. The study results revealed that the data sizes effect model performances up to a certain level.
Citation - WoS: 4
Citation - Scopus: 5
Automata Networks as Preprocessing Technique of Artificial Neural Network in Estimating Primary Production and Dominating Phytoplankton Levels in a Reservoir
(Elsevier, 2006) Kilic, Hurevren; Soyupak, Selcuk; Gurbuz, Hasan; Kivrak, Ersin; Computer Engineering; 06. School Of Engineering; 01. Atılım University
Artificial Neural Networks (ANN) is computational architectures that can be used for estimating primary production levels and dominating phytoplankton species in reservoirs. Automata Networks (AN) were applied as a pre-processing method with subsequent ANN model development for Demirdoven Dam Reservoir. The primary purpose of using preprocessing technique was to distinguish the suitable and appropriate constituents of the input parameters' matrix, to eliminate redundancy, to enhance prediction power and calculation efficiency. The data were collected monthly over two years. The applications have yielded following results: The correlation coefficients (r values) between predicted and observed counts were as high as 0.83, 0.87, 0.83 and 0.88 for Cyclotella ocellata, Sphaerocystis schroeteri, Staurastrum longiradiatum counts, and Chlorophyll-a (Chl-a) concentrations respectively with AN. The performance of AN based pre-processing technique was compared with the performance of a well-known pre-processing technique, namely Principle Component Analysis(PCA), experimentally. r values between the predicted and observed C. ocellata, S. schroeteri and S. longiradiatum counts, and (Chl-a) were as high as 0.80, 0.86, 0.81 and 0.86 respectively with PCA. (c) 2006 Elsevier B.V. All rights reserved.
Citation - WoS: 4
Citation - Scopus: 6
An Automata Networks Based Preprocessing Technique for Artificial Neural Network Modelling of Primary Production Levels in Reservoirs
(Elsevier, 2007) Kilic, Hurevren; Soyupak, Selcuk; Tuzun, Ilhami; Ince, Ozlem; Basaran, Gokben; Computer Engineering; 06. School Of Engineering; 01. Atılım University
Primary production in lakes and reservoirs develops as a result of complex reactions and interactions. Artificial neural networks (ANN) emerges as an approach in quantification of primary productivity in reservoirs. Almost all of the past ANN applications employed input data matrices whose vectors represent either water quality parameters or environmental characteristics. Most of the time, the components of input matrices are determined using expert opinion that implies possible factors that affect output vector. Major disadvantage of this approach is the possibility of ending-up with an input matrix that may have high correlations between some of its components. In this paper, an automata networks (AN) based preprocessing technique was developed to select suitable and appropriate constituents of input matrix to eliminate redundancy and to enhance calculation efficiency. The proposed technique specifically provides an apriori rough behavioral modeling through identification of minimal AN interaction topology. Predictive ANN models of primary production levels were developed for a reservoir following AN based pre-modeling step. The achieved levels of model precisions and performances were acceptable: the calculated root mean square error values (RMSE) were low; a correlation coefficient (R) as high as 0.83 was achieved with an ANN model of a specific structure. (c) 2006 Elsevier B.V. All rights reserved.
A Comparison of Regression, Neural Network and Fuzzy Logic Models for Estimating Chlorophyll-A Concentrations in Reservoirs
(Centre Environment Social & Economic Research Publ-ceser, 2005) Chen, Ding-Geng; Soyupak, Selcuk; 01. Atılım University
A comparison is conducted in this paper for the multiple linear regression, neural network and fuzzy logic models for their ability to estimate pseudo steady state chlorophyll-a concentrations in a very large and deep dam reservoir that exhibits high spatial and temporal variability. The utilized data set include chlorophyll-a concentrations as an indicator of primary productivity as well as several other water quality variables such as alkalinity, PO4 phosphorus, water temperature and dissolved oxygen concentrations as independent environmental variables. Using the conventional model criteria of correlation coefficient and mean square errors, the fuzzy logic model performed the best with the neural network model better than multiple linear regression model.