This work is motivated by the interest in feature selection that greatly affects the detection accuracy of a classifier. The goals of this paper are (i) identifying optimal feature subset using a novel wrapper based feature selection algorithm called Shapley Value Embedded Genetic Algorithm (SVEGA), (ii) showing the improvement in the detection accuracy of the Artificial Neural Network (ANN) classifier with the optimal features selected, (iii) evaluating the performance of proposed SVEGA-ANN model on the medical datasets. The medical diagnosis system has been built using a wrapper based feature selection algorithm that attempts to maximize the specificity and sensitivity (in turn the accuracy) as well as by employing an ANN for classification. Two memetic operators namely include and remove features (or genes) are introduced to realize the genetic algorithm (GA) solution. The use of GA for feature selection facilitates quick improvement in the solution through a fine tune search. An extensive experimental evaluation of the proposed SVEGA-ANN method on 26 benchmark datasets from UCI Machine Learning repository and Kent ridge repository, with three conventional classifiers, outperforms state-of-the-art systems in terms of classification accuracy, number of selected features and running time.
This paper presents a hybrid model based on the displacement back analysis to estimate the earth stress magnitude and direction from the obtained borehole displacements. An artificial neural network (ANN) is used to map the non-linear relationship between the maximum horizontal earth stress, σH, the minimum horizontal earth stress, σh, the direction of the largest horizontal earth stress, θ and the borehole displacements. The genetic algorithm (GA) is used to search the set of unknown earth stresses and direction according to the objective function. Results of the numerical experiments show that the displacement back analysis method can effectively identify the earth stress based on the wellbore motions during drilling. and Obsahuje seznam literatury
This paper presents a semi-global mathematical model for an analysis of a signal of amperometric biosensors. Artificial neural networks were applied to an analysis of the biosensor response to multi-component mixtures. A large amount of the learning and test data was synthesized using computer simulation of the biosensor response. The biosensor signal was analyzed with respect to the concentration of each component of the mixture. The paradigm of locally weighted linear regression was used for retraining the neural networks. The application of locally weighted regression significantly improved the quality of the prediction of the concentrations.
This paper presents an empirical model and a three-layer (7:11:1) artificial neural network (ANN) approach for the determination of completely mixed activated sludge reactor volume (CMASRV). CMASRV values were estimated by a new mathematical formulation and a three-layer ANN model for 1,000 different artificial scenarios given in a wide range of seven biological variables. The predicted results obtained from each stochastic approach were compared with the well-known steady state volume model based on mass balance equations. The computational analysis showed that the proposed empirical model and ANN outputs were obviously in agreement with the steady-state volume model and all the predictions proved to be satisfactory with a correlation coefficient of about 0.9989 and 1, respectively. The maximum volume deviations from the steady-state volume equation were recorded as only 7.17% and 6.89% for the proposed model and ANN outputs respectively. In addition to volume comparison, waste sludge mass flow rates (PX), food to mass ratios (F/M), hydraulic retention times (HRTs), volumetric organic loads (LV) and oxygen requirements (ORs) were also compared for each model, and significant points of proposed approaches were evaluated.
The prediction of traffic volume over time is very important to control the flow of traffic on a road network. Traffic count is usually averaged over time to predict for the larger time domain. This paper aims at finding the detail variation of a systematic survey of hourly traffic volume data over a time of four years along the North Bengal corridor of Bangladesh (at Jamuna toll collection point) and its equivalent numerical model by using a Artificial Neural Network. The Neural Network is trained with the intermittent data of 13 weeks over four years and the missing data is interpreted with quite reasonable accuracy (12.67% MAE) with this ANN model. The ANN model captured the variety of trends of the traffic data very accurately as has been depicted in the paper
Gears are used to transmit power and motion in mechanical, electrical and chemical process industries. Influenced by vibration, torque, temperature, lubrication & specific film thickness, the gear teeth contacts may experience change leading to unexpected failures such as wear, scutting, pitting and micro-pitting on teeth surface. In order to avoid these damages, continuous monitoring is essential using knowledge based systems, Generic capability of artificial neural network (ANN) is exploited to formulate prediction and classification based on heuristic models of condition of lubricating oil in spur gears. Based on the loading conditions such as vibration, temperature and torque, the algorithm predicts film thickness to classify oil conditions as elastohydrodynamic (EHD), mixed wear and severe wear that helps in detecting faults in gear operation. and Obsahuje seznam literatury
Uniaxial compressive strength (UCS) represents one of the key mechanical properties used to characterize rocks along with the other important properties of porosity and density. While several studies have proved the accuracy of artificial intelligence in modeling UCS, some authors believe that the use of artificial intelligence is not practical in predicting. The present paper highlights the ability of an artificial neural network (ANN) as an accurate and revolutionary method with regression models, as a conventional statistical analysis, to predict UCS within carbonate rocks and mortar. Thus, ANN and multiple linear regressions (MLR) were applied to estimate the UCS values of the tested samples. For experimentation we carried out ultrasonic measurements on cubic samples before testing uniaxial compressive strength perpendicularly to the stress direction. The models were performed to correlate effective porosity, density and ultrasonic velocity to the UCS measurements. The resulting models would allow the prediction of carbonate rocks and mortar’s UCS values usually determined by laborious experiments. Although the results demonstrate the usefulness of the MLP method as a simple, practical and economical model, the ANN model is more accurate.
This research investigated the effect of different drought conditions on Barley (Hordeum vulgare L.) yield in North Dakota, USA, using Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) methods. Though MLR method is widely used, the ANN method has not been used in the past to investigate the effect of droughts on barley yields to the best of authors knowledge. It is found from this study that the ANN model performs better than MLR in estimating barley yield. In this paper, the ANN is proposed as a viable alternative method or in combination with MLR to investigate the impact of droughts on crop yields.
This paper analyses rent-based determinants of earthquake damage from an urban planning perspective with the data gathered from Adapazari, Turkey, after the disaster in 1999 Eastern Marmara Earthquake (EME). The study employs linear regression, log-linear regression, and artificial neural networks (ANN) methods for cross-verification of results and for finding out the significant urban rent attribute(s) responsible for the damage. All models used are equally capable of predicting the earthquake damage and converge to similar results even if the data are limited. Of the rent variables, the physical density is proved to be especially significant in predicting earthquake damage, while the land value contributes to building resistance. Thus, urban rent can be the primary tool for planners to help reduce the fatalities in preventive planning studies.
The Internet has become an important source of information for physicians seeking immediate data for the management of patients and for those developing decision-making methodologies and guidelines for clinical practice. In this study, components and subsystems of a medical decision support system are presented. An artificial neural network model, which is one of the subsystems of the differential diagnosis component, has been proposed as a reasoning tool to support medical diagnosis. The input data of artificial neural network models used in different medical diagnosis can be obtained via the Internet. The present study is concerned with the application of artificial neural network model to diabetes prediction. Demographic and medical data of diabetics and non-diabetics obtained via the Internet were used as the artificial neural network inputs. The accuracy of the neural network's results has shown that the diabetes prediction is feasible by the neural network described in this study.