Hybrid modified Cuckoo Search-Neural Network in chronic kidney disease classification
01 Jun 2017-pp 164-167
TL;DR: The experimental results depicted that the NN-MCS has the ability to detect CKD more efficiently compared to any other existing model.
Abstract: Chronic kidney failure (chronic kidney disease ‘CKD’) is a serious disease that related to the gradual loss of kidney function. It is considered one of the health threats in the developing and undeveloped countries At early stages, few symptoms can be detected, where the CKD may not become obvious until significant kidney function impaired occur. CKD treatment focuses on reducing the kidney damage progression by controlling the underlying cause, which requires disease detection at initial stages. In early addition, the financial burden of the treatment and future consequences of CKD requires early detection. In the present work a modified Cuckoo Search (MCS) trained Neural Network (NN) or NN-MCS based model is proposed to detect CKD. The NN-MCS model has been proposed to overcome the problem of using local search based learning algorithms to train the NNs. The NN weight vector is optimized by applying MCS for NN training. A comparative study with eminent classifiers, namely the Multilayer Perceptron Feed-forward Network (MLP-FFN) and NN based on Particle Swarm Optimization (PSO-NN). The classifiers performance is measured in terms of different performance metrics. The experimental results depicted that the NN-MCS has the ability to detect CKD more efficiently compared to any other existing model.
Citations
More filters
TL;DR: Comparing the D-ACO algorithm with existing methods, the presented intelligent system outperformed the other methodologies with a significant improvisation in classification accuracy using fewer features.
Abstract: At present times, healthcare systems are updated with advanced capabilities like machine learning (ML), data mining and artificial intelligence to offer human with more intelligent and expert healthcare services. This paper introduces an intelligent prediction and classification system for healthcare, namely Density based Feature Selection (DFS) with Ant Colony based Optimization (D-ACO) algorithm for chronic kidney disease (CKD). The proposed intelligent system eliminates irrelevant or redundant features by DFS in prior to the ACO based classifier construction. The proposed D-ACO framework three phases namely preprocessing, Feature Selection (FS) and classification. Furthermore, the D-ACO algorithm is tested using benchmark CKD dataset and the performance are investigated based on different evaluation factors. Comparing the D-ACO algorithm with existing methods, the presented intelligent system outperformed the other methodologies with a significant improvisation in classification accuracy using fewer features.
118 citations
TL;DR: This article describes how howSequences areﻷ attributedﻴtemporal�characteristicsﻵeitherﻰ�explicitlyﻡ�orﻢimplicitly £2.5m Cybersecurity.
Abstract: This article describes how sequential data modeling is a relevant task in Cybersecurity. Sequences are attributed temporal characteristics either explicitly or implicitly. Recurrent neural networks (RNNs) are a subset of artificial neural networks (ANNs) which have appeared as a powerful, principle approach to learn dynamic temporal behaviors in an arbitrary length of large-scale sequence data. Furthermore, stacked recurrent neural networks (S-RNNs) have the potential to learn complex temporal behaviors quickly, including sparse representations. To leverage this, the authors model network traffic as a time series, particularly transmission control protocol / internet protocol (TCP/IP) packets in a predefined time range with a supervised learning method, using millions of known good and bad network connections. To find out the best architecture, the authors complete a comprehensive review of various RNN architectures with its network parameters and network structures. Ideally, as a test bed, they use the existing benchmark Defense Advanced Research Projects Agency / Knowledge Discovery and Data Mining (DARPA) / (KDD) Cup ‘99’ intrusion detection (ID) contest data set to show the efficacy of these various RNN architectures. All the experiments of deep learning architectures are run up to 1000 epochs with a learning rate in the range [0.01-0.5] on a GPU-enabled TensorFlow and experiments of traditional machine learning algorithms are done using Scikit-learn. Experiments of families of RNN architecture achieved a low false positive rate in comparison to the traditional machine learning classifiers. The primary reason is that RNN architectures are able to store information for long-term dependencies over time-lags and to adjust with successive connection sequence information. In addition, the effectiveness of RNN architectures are shown for the UNSW-NB15 data set. KEywoRDS Deep Learning (DL) Approaches, Gated Recurrent Unit (GRU), Intrusion Detection (ID) Data Sets, KDDCup ’99’, Long Short-Term Memory (LSTM), Machine Learning (ML), Recurrent Neural Network (RNN), UNSW-NB15
59 citations
TL;DR: A new intrusion detection model is created, which is able to classify the binary-class, triple- class, and multi-class cyber-attacks and power-system incidents and compares the proposed model with other commonly used classifiers.
Abstract: The smart grid is a revolutionary, intelligent, next-generation power system. Due to its cyber infrastructure nature, it must be able to accurately and detect potential cyber-attacks and take appropriate actions in a timely manner. This paper creates a new intrusion detection model, which is able to classify the binary-class, triple-class, and multi-class cyber-attacks and power-system incidents. The intrusion detection model is based on a whale optimization algorithm (WOA)-trained artificial neural network (ANN). The WOA is applied to initialize and adjust the weight vector of the ANN to achieve the minimum mean square error. The proposed WOA-ANN model can address the challenges of attacks, failure prediction, and failure detection in a power system. We utilize the Mississippi State University and Oak Ridge National Laboratory databases of power-system attacks to demonstrate the proposed model and show the experimental results. The WOA is able to train the ANN to find the optimal weights. We compare the proposed model with other commonly used classifiers. The comparison results show the superiority of the proposed WOA-ANN model.
59 citations
01 Jan 2019
TL;DR: This proposed method is based on deep neural network which predicts the presence or absence of chronic kidney disease with an accuracy of 97% and shows better results which is implemented using the cross-validation technique to keep the model safe from overfitting.
Abstract: The progression of the chronic kidney disease and methodologies to diagnose chronic kidney disease is a challenging problem which can reduce the cost of treatment. We studied 224 records of chronic kidney disease available on the UCI machine learning repository named chronic kidney diseases dating back to 2015. Our proposed method is based on deep neural network which predicts the presence or absence of chronic kidney disease with an accuracy of 97%. Compared to other available algorithms, the model we built shows better results which is implemented using the cross-validation technique to keep the model safe from overfitting. This automatic chronic kidney disease treatment helps reduce the kidney damage progression, but for this chronic kidney disease detection at initial stage is necessary.
36 citations
TL;DR: These methods can assist specialists in determining the stage of chronic renal disease and enhance the precision of medical diagnostics used to diagnose illnesses.
Abstract: Kidney failure occurs whenever the kidney stops to operate properly and would be unable to cleanse or refine the bloodstream as it should. Chronic kidney disease (CKD) is a potentially fatal consequence. If this condition is diagnosed early, its progression can be delayed. There are various factors that increase the likelihood of developing kidney failure. As a consequence, in order to detect this potentially fatal condition early on, these risk factors must be checked on a regular basis before the individual's health deteriorates. Furthermore, it lowers the cost of therapy. The chronic kidney or renal disease will be recognized in this work utilizing fuzzy and adaptive neural fuzzy inference systems. The fundamental purpose of this initiative is to enhance the precision of medical diagnostics used to diagnose illnesses. Nephron functioning, glucose levels, systolic and diastolic blood pressure, maturity level, weight and height, and smoking are all elements to consider while developing a fuzzy and adaptable neural fuzzy inference system. The output variable describes a specific patient's stage of chronic renal disease based on input factors such as stage 1, stage 2, stage 3, stage 4, and stage 5. The outcome will show the present stage of a patient's kidney. As a result, these methods can assist specialists in determining the stage of chronic renal disease. MATLAB software is used to create the fuzzy and neural fuzzy inference systems.
30 citations
References
More filters
01 Mar 1996
TL;DR: The article discusses the motivations behind the development of ANNs and describes the basic biological neuron and the artificial computational model, and outlines network architectures and learning processes, and presents some of the most commonly used ANN models.
Abstract: Artificial neural nets (ANNs) are massively parallel systems with large numbers of interconnected simple processors. The article discusses the motivations behind the development of ANNs and describes the basic biological neuron and the artificial computational model. It outlines network architectures and learning processes, and presents some of the most commonly used ANN models. It concludes with character recognition, a successful ANN application.
4,281 citations
"Hybrid modified Cuckoo Search-Neura..." refers background in this paper
...Regardless of the initial weights, for finite iterations, it is required to accomplish optimal weights using the perceptron learning [12]....
[...]
...However, apart from its immense success in several real life scenarios [11-13], the model faced a severe problem of getting stuck in local optima while searching for optimal weights for NN....
[...]
...ANNs are considered robust machine learning approaches capable of approximating complex patterns [11, 12]....
[...]
TL;DR: Experiments show that SCG is considerably faster than BP, CGL, and BFGS, and avoids a time consuming line search.
3,882 citations
"Hybrid modified Cuckoo Search-Neura..." refers background or methods in this paper
...The present NN-MCS performance is compared to MLP-FFN trained with scaled conjugate descent learning procedure [13] and the back-propagation NN....
[...]
...However, apart from its immense success in several real life scenarios [11-13], the model faced a severe problem of getting stuck in local optima while searching for optimal weights for NN....
[...]
TL;DR: The data suggest that efforts to reduce mortality in this population should be focused on treatment and prevention of coronary artery disease, congestive heart failure, diabetes mellitus, and anemia.
Abstract: Background Chronic kidney disease is the primary cause of end-stage renal disease in the United States. The purpose of this study was to understand the natural history of chronic kidney disease with regard to progression to renal replacement therapy (transplant or dialysis) and death in a representative patient population. Methods In 1996 we identified 27 998 patients in our health plan who had estimated glomerular filtration rates of less than 90 mL/min per 1.73 m 2 on 2 separate measurements at least 90 days apart. We followed up patients from the index date of the first glomerular filtration rates of less than 90 mL/min per 1.73 m 2 until renal replacement therapy, death, disenrollment from the health plan, or June 30, 2001. We extracted from the computerized medical records the prevalence of the following comorbidities at the index date and end point: hypertension, diabetes mellitus, coronary artery disease, congestive heart failure, hyperlipidemia, and renal anemia. Results Our data showed that the rate of renal replacement therapy over the 5-year observation period was 1.1%, 1.3%, and 19.9%, respectively, for the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (K/DOQI) stages 2, 3, and 4, but that the mortality rate was 19.5%, 24.3%, and 45.7%. Thus, death was far more common than dialysis at all stages. In addition, congestive heart failure, coronary artery disease, diabetes, and anemia were more prevalent in the patients who died but hypertension prevalence was similar across all stages. Conclusion Our data suggest that efforts to reduce mortality in this population should be focused on treatment and prevention of coronary artery disease, congestive heart failure, diabetes mellitus, and anemia.
1,580 citations
Additional excerpts
...The problem has become more challenging in absence of a well-accepted method for predicting CKD, which in future can lead to an end-stage renal failure (ESRD) [2, 3]....
[...]
TL;DR: This paper presents a more extensive comparison study using some standard test functions and newly designed stochastic test functions to apply the CS algorithm to solve engineering design optimisation problems, including the design of springs and welded beam structures.
Abstract: A new metaheuristic optimisation algorithm, called cuckoo search (CS), was developed recently by Yang and Deb (2009). This paper presents a more extensive comparison study using some standard test functions and newly designed stochastic test functions. We then apply the CS algorithm to solve engineering design optimisation problems, including the design of springs and welded beam structures. The optimal solutions obtained by CS are far better than the best solutions obtained by an efficient particle swarm optimiser. We will discuss the unique search features used in CS and the implications for further research.
1,339 citations
"Hybrid modified Cuckoo Search-Neura..." refers background in this paper
...Three descriptive CS algorithm rules are as follows [14]: • One egg is laid by each cuckoo at a time with deposits it in an arbitrary nest; • The superior nests containing high quality solutions (eggs) will remain to the next iteration; • The accessible fixed number of host nests is assumed....
[...]
...Cuckoo search is considered the most efficient meta-heuristic optimization procedure [14], which driven by obligates brood parasitism nature of cuckoo birds....
[...]
TL;DR: In this article, a nonlinear transformation of two independent uniform random variables into one stable random variable is presented, which is a continuous function of each of the uniform random variable, and of α and a modified skewness parameter β' throughout their respective permissible ranges.
Abstract: A new algorithm is presented for simulating stable random variables on a digital computer for arbitrary characteristic exponent α(0 < α ≤ 2) and skewness parameter β(-1 ≤ β ≤ 1). The algorithm involves a nonlinear transformation of two independent uniform random variables into one stable random variable. This stable random variable is a continuous function of each of the uniform random variables, and of α and a modified skewness parameter β' throughout their respective permissible ranges.
1,124 citations
"Hybrid modified Cuckoo Search-Neura..." refers methods in this paper
...Modeling the Lévy flight in the CS algorithm has been applied in [24]....
[...]
...Therefore, a faster method proposed in [24] that can be used in some real world problems....
[...]