In many cases, the Neural Networks (NN) method outperforms standard machine learning techniques. The resource learning architecture can be enhanced to boost its performance even more. The neural networks [25,26,27,28,29] models have been used for the detection of kidney disease. The majority of currently available CKD models have a low classification accuracy. As a result, this research introduces a novel model for Chronic Kidney Disease.
Neural Networks Rajasekaran Pdf 62
Download Zip: https://8novacxcespe.blogspot.com/?file=2vKb5U
Krishnamurthy S. et al. [38] developed various artificial intelligence models to predict Chronic Kidney Disease. The LightGBM model selected the most important features for CKD prediction: age, gout, diabetes mellitus, use of sulfonamides, and angiotensins. The convolutional neural networks achieved the best performance and the highest AUROC metric, 0.954, compared to other models. Mohamed Elhoseny et al. [19] presented an intelligent prediction system for Chronic Kidney Disease. The density-based Feature Selection method eliminates the irrelevant features and then passes selected features to the Ant Colony-based Optimization classifier to predict CKD. Singh and Jain [39] presented novel hybrid approach for diagnose CKD and achieved 92.5 % of prediction accuracy. An artificial neural network for CKD diagnosis was proposed by Neves et al. [25]. The diagnostic sensitivity values ranged from 93.1% to 94.9%, and the diagnostic specificity values ranged from 91.9% to 94.2% in this work.
Vasquez-Morales et al. [27] used large CKD data to generate a neural network classifier, and the model was 95% accurate. Makino et al. [28] used textual data to extract patients diagnoses and treatment information in order to forecast the course of diabetic kidney disease. Ren et al. [29] developed a predictive model for the identification of CKD from an Electronic Health Records (EHR) data set. This proposed model is based on a neural network framework that encodes and decodes the textual and numerical information from EHR. Ma F. et al. [40] develop a deep neural network model to detect chronic renal disease. The presented model obtained the highest accuracy compared to ANN and SVM. Almansour et al. [41] devised a way for preventing CKD using machine learning. The SVM and ANN were among the machine learning classification algorithms used by the researchers. The results of the experiments revealed that ANN had a greater accuracy of 99.75% than SVM.
J. Qin et al. [42] presented a machine learning method for the early detection of CKD. They used logistic regression, random forest, SVM, naive Bayes classifier, KNN, and feed-forward neural network to develop their models. The most accurate classification model was random forest, which had a 99.75% accuracy rate. Z. Segal et al. [43] presented a machine learning technique based on an ensemble tree (XGBoost) for the early diagnosis of renal illness. The presented model was compared against Random Forest, CatBoost, Regression with Regularization. The proposed model showed better performance in all matrices, including c-statistics 0.93, sensitivity 0.715, and specificity 0.958. Khamparia et al. [44] developed a deep learning model for early detection of CKD, in which features were selected from multimedia data using a stacked autoencoder model. The authors used A SoftMax classifier to predict the final class. It was observed that the proposed model achieved the highest performance in comparison to conventional classification techniques on the UCI CKD data set.
This paper deals on artificial intelligence (AI) application for the estimation of kerf geometry and hole diameters for laser micro-cutting and laser micro-drilling operations. To this aim laser cutting and laser drilling operation were performed on NIMONIC 263 superalloy sheet, 0.38 mm in nominal thickness, by way of a 100 W fibre laser in modulated wave regime. Linear cuts and holes (by trepanning) were performed fixing the average power at 80 W and changing the pulse duration, the cutting speed, the focus depth and the laser path (the latter only for the drilling operations). Kerf width and the holed diameter, at the upper and downsides, were measured by digital microscopy. Different artificial neural networks (ANNs) were developed and tested to predict the kerf widths and the diameters (at the upper and downside). Two ANNs were addressed to the linear cutting process modelling; also, two further ANNs were developed for micro-drilling on the base of the linear cutting process features. The networks were trained with a subset of data containing the process conditions and the kerf/hole geometry. The ANN test was performed with the remaining data. The results show that ANNs can model the cut and hole geometry as a function of the process parameters. Moreover, the ANN trained with kerf geometry is more efficient. Therefore, a functional correlation between the kerf geometries achievable in the linear cutting process and micro-drilling was assessed.
As stated before, 4 ANNs were developed for the prediction of Uk, Dk, Din and Dout. They are indicated in this paper as, respectively, Uk-Net, Dk-Net, Din-Net and Dout-Net. In particular, backpropagation feedforward neural networks were implemented in MATLAB environment and trained with the experimental data achieved. During the training, several configurations were considered. The variables involved were the main structure parameters such as the number of hidden layers, number of neurons, transfer functions and training algorithm. No optimization was involved in network size assessment, which was determined through a trial and error process evaluating the means square error (MSE), as reported in Eq. 1.
For the sake of briefness, only the networks offering the best prediction, which structures and training parameters are resumed in Table 4, are reported and discussed. The network performances were accounted by means of three indicators, which are mean percentage error xm (Eq. 2), standard deviation xdv.std. (Eq. 3) and linear least square regression.
Although the heat diffusion in the linear configuration is less tangled than in micro-drilling, due to the simple geometry of the cut, the ANNs are more complex. Indeed, these networks present a higher number of hidden layers and neurons. However, the micro-drilling ANNs have 1 more input neuron that refers to the linear cut geometry. It is a critical parameter and a precious aid for achieving predictions about the kerf characterization. Also, this confirms how the linear kerf may be a valuable starting point for a micro-drilling kerf characterization. 2ff7e9595c
Comments