Forest Type Classification: A Hybrid NN-GA Model Based Approach
01 Jan 2016-pp 227-236
TL;DR: The authors have proposed a GA trained Neural Network classifier to tackle the task of classify tree species and one mixed forest class using geographically weighted variables calculated for Cryptomeria japonica and Chamaecyparis obtusa.
Abstract: Recent researches have used geographically weighted variables calculated for two tree species, Cryptomeria japonica (Sugi, or Japanese Cedar) and Chamaecyparis obtusa (Hinoki, or Japanese Cypress) to classify the two species and one mixed forest class. In machine learning context it has been found to be difficult to predict that a pixel belongs to a specific class in a heterogeneous landscape image, especially in forest images, as ground features of nearly located pixel of different classes have very similar spectral characteristics. In the present work the authors have proposed a GA trained Neural Network classifier to tackle the task. The local search based traditional weight optimization algorithms may get trapped in local optima and may be poor in training the network. NN trained with GA (NN-GA) overcomes the problem by gradually optimizing the input weight vector of the NN. The performance of NN-GA has been compared with NN, SVM and Random Forest classifiers in terms of performance measures like accuracy, precision, recall, F-Measure and Kappa Statistic. The results have been found to be satisfactory and a reasonable improvement has been made over the existing performances in the literature by using NN-GA.
Citations
More filters
TL;DR: A novel application of Particle Swarm Optimization (PSO) trained Artificial Neural Network (ANN) has been employed to separate the patients having Dengue fevers from those who are recovering from it or do not have DF.
Abstract: A mosquito borne pathogen called Dengue virus (DENV) has been emerged as one of the most fatal threats in the recent time. Infections can be in two main forms, namely the DF (Dengue Fever), and DHF (Dengue Hemorrhagic Fever). An efficient detection method for both fever types turns out to be a significant task. Thus, in the present work, a novel application of Particle Swarm Optimization (PSO) trained Artificial Neural Network (ANN) has been employed to separate the patients having Dengue fevers from those who are recovering from it or do not have DF. The ANN’s input weight vector are optimized using PSO to achieve the expected accuracy and to avoid premature convergence toward the local optima. Therefore, a gene expression data (GDS5093 dataset) available publicly is used. The dataset contains gene expression data for DF, DHF, convalescent and healthy control patients of total 56 subjects. Greedy forward selection method has been applied to select most promising genes to identify the DF, DHF and normal (either convalescent or healthy controlled) patients. The proposed system performance was compared to the multilayer perceptron feed-forward neural network (MLP-FFN) classifier. Results proved the dominance of the proposed method with achieved accuracy of 90.91 %.
56 citations
01 Jan 2018
50 citations
TL;DR: A modified bag-of-features method has been proposed to select the most promising genes in the classification process and results indicated a highly statistically significant improvement with the proposed classifier over the traditional ANN-CS model.
Abstract: Dengue fever detection and classification have a vital role due to the recent outbreaks of different kinds of dengue fever. Recently, the advancement in the microarray technology can be employed for such classification process. Several studies have established that the gene selection phase takes a significant role in the classifier performance. Subsequently, the current study focused on detecting two different variations, namely, dengue fever (DF) and dengue hemorrhagic fever (DHF). A modified bag-of-features method has been proposed to select the most promising genes in the classification process. Afterward, a modified cuckoo search optimization algorithm has been engaged to support the artificial neural (ANN-MCS) to classify the unknown subjects into three different classes namely, DF, DHF, and another class containing convalescent and normal cases. The proposed method has been compared with other three well-known classifiers, namely, multilayer perceptron feed-forward network (MLP-FFN), artificial neural network (ANN) trained with cuckoo search (ANN-CS), and ANN trained with PSO (ANN-PSO). Experiments have been carried out with different number of clusters for the initial bag-of-features-based feature selection phase. After obtaining the reduced dataset, the hybrid ANN-MCS model has been employed for the classification process. The results have been compared in terms of the confusion matrix-based performance measuring metrics. The experimental results indicated a highly statistically significant improvement with the proposed classifier over the traditional ANN-CS model.
43 citations
Cites background or methods from "Forest Type Classification: A Hybri..."
...Previously, several recently developed meta-heuristic algorithms have been employed to train ANNs apart from traditional ones, such as genetic algorithm [40], particle swarm optimization [32], and the non-dominated sorting genetic algorithm [41]....
[...]
...ter quality prediction [36], electrical energy output prediction [33], dengue fever classification [42], and environmental application [40]....
[...]
01 Aug 2017
TL;DR: Experimental results clearly show the superiority of the proposed NN-NSGA-II model with different features, which has been evaluated using various performances measuring metrics such as accuracy, precision, recall and F-measure.
Abstract: Automated, efficient and accurate classification of skin diseases using digital images of skin is very important for bio-medical image analysis. Various techniques have already been developed by many researchers. In this work, a technique based on meta-heuristic supported artificial neural network has been proposed to classify images. Here 3 common skin diseases have been considered namely angioma, basal cell carcinoma and lentigo simplex. Images have been obtained from International Skin Imaging Collaboration (ISIC) dataset. A popular multi objective optimization method called Non-dominated Sorting Genetic Algorithm — II is employed to train the ANN (NNNSGA-II). Different feature have been extracted to train the classifier. A comparison has been made with the proposed model and two other popular meta-heuristic based classifier namely NN-PSO (ANN trained with Particle Swarm Optimization) and NN-GA (ANN trained with Genetic algorithm). The results have been evaluated using various performances measuring metrics such as accuracy, precision, recall and F-measure. Experimental results clearly show the superiority of the proposed NN-NSGA-II model with different features.
39 citations
Cites background from "Forest Type Classification: A Hybri..."
...Early stage detection of the disease type is important since appropriate treatments can be applied based on the type of the disease [3]....
[...]
01 Oct 2017
TL;DR: Experimental results indicated towards the superiority of the proposed bag-of-features enabled NN-NSGA-II model in terms of testing phase confusion matrix based performance measuring metrics.
Abstract: The current work proposes a neural based detection method of two different skin diseases using skin imaging. Skin images of two diseases namely Basel Cell Carcinoma and Skin Angioma are utilized. SIFT feature extractor has been employed followed by a clustering phase on feature space in order to reduce the number of features suitable for neural based models. The extracted bag-of-features modified dataset is used to train metaheuristic supported hybrid Artificial Neural Networks to classify the skin images in order to detect the diseases under study. A well-known multi objective optimization technique called Non-dominated Sorting Genetic Algorithm — II is used to train the ANN (NN-NSGA-II). The proposed model is further compared with two other well-known metaheuristic based classifier namely NN-PSO (ANN trained with PSO) and NN-CS (ANN trained with Cuckoo Search) in terms of testing phase confusion matrix based performance measuring metrics such as accuracy, precision, recall and F-measure. Experimental results indicated towards the superiority of the proposed bag-of-features enabled NN-NSGA-II model.
36 citations
References
More filters
TL;DR: It is established the functional equivalence of a generalized class of Gaussian radial basis function networks and the full Takagi-Sugeno model (1983) of fuzzy inference and the more general framework allows the removal of some of the restrictive conditions.
Abstract: We establish the functional equivalence of a generalized class of Gaussian radial basis function (RBFs) networks and the full Takagi-Sugeno model (1983) of fuzzy inference. This generalizes an existing result which applies to the standard Gaussian RBF network and a restricted form of the Takagi-Sugeno fuzzy system. The more general framework allows the removal of some of the restrictive conditions of the previous result.
164 citations
TL;DR: Results are presented to show that this two-stage process leads to separation of corn and soybean, and of several minor classes that would otherwise be overwhelmed in any practical one-stage clustering.
Abstract: In this paper, a segmentation procedure that utilizes a clustering algorithm based upon fuzzy set theory is developed. The procedure operates in a nonparametric unsupervised mode. The feasibility of the methodology is demonstrated by segmenting a six-band Landsat-4 digital image with 324 scan lines and 392 pixels per scan line. For this image, 100-percent ground cover information is available for estimating the quality of segmentation. About 80 percent of the imaged area contains corn and soybean fields near the peak of their growing season. The remaining 20 percent of the image contains 12 different types of ground cover classes that appear in regions of diffferent sizes and shapes. The segmentation method uses the fuzzy c-means algorithm in two stages. The large number of clusters resulting from this segmentation process are then merged by use of a similarity measure on the cluster centers. Results are presented to show that this two-stage process leads to separation of corn and soybean, and of several minor classes that would otherwise be overwhelmed in any practical one-stage clustering.
145 citations
TL;DR: An implementation of an artificial neural network (ANN) which performs unsupervised detection of recognition categories from arbitrary sequences of multivalued input patterns called SARTNN, which gives good results in terms of ease of use, parameter robustness and computation time.
Abstract: This article presents an implementation of an artificial neural network (ANN) which performs unsupervised detection of recognition categories from arbitrary sequences of multivalued input patterns. The proposed ANN is called Simplified Adaptive Resonance Theory Neural Network (SARTNN). First, an Improved Adaptive Resonance Theory 1 (IARTl)-based neural network for binary pattern analysis is discussed and a Simplified ARTl (SART1) model is proposed. Second, the SARTl model is extended to multivalued input pattern clustering and SART” is presented. A normalized coefficient which measures the degree of match between two multivalued vectors, the Vector Degree of Match (VDM), provides SARTNN with the metric needed to perform clustering. Every ART architecture guarantees both plasticity and stability to the unsupervised learning stage. The SARTNN plasticity requirement is satisfied by implementing its attentional subsystem as a self-organized, feed-forward, flat Kohonen’s ANN (KANN). The SARTNN stability requirement is properly driven by its orienting subsystem. SARTNN processes multivalued input vectors while featuring a simplified architectural and mathematical model with respect to both the ARTl and the AkT2 models, the latter being the ART model fitted to multivalued input pattern categorization. While the ART2 model exploits ten user-defined parameters, SARTNN requires only two user-defined parameters to be run: the first parameter is the vigilance threshold, p, that affects the network’s sensibility in detecting new output categories, whereas the second parameter, T, is related to the network's learning rate. Both parameters have an intuitive physical meaning and allow the user to choose easily the proper discriminating power of the category extraction algorithm. The SARTNN performance is tested as a satellite image clustering algorithm. A chromatic component extractor is recommended in a satellite image preprocessing stage, in order to pursue SARTNN invariant pattern recognition. In comparison with classical clustering algorithms like ISODATA, the implemented system gives good results in terms of ease of use, parameter robustness and computation time. SARTNN should improve the performance of a Constraint Satisfaction Neural Network (CSNN) for image segmentation. SARTNN, exploited as a self-organizing first layer, should also improve the performance of both the Counter Propagation Neural Network (CPNN) and the Reduced connectivity Coulomb Energy Neural Network (RCENN).
110 citations
TL;DR: The concept of chromosome differentiation, commonly witnessed in nature as male and female sexes, is incorporated in genetic algorithms with variable length strings for designing a nonparametric classification methodology and its significance in partitioning different landcover regions from satellite images is demonstrated.
Abstract: The concept of chromosome differentiation, commonly witnessed in nature as male and female sexes, is incorporated in genetic algorithms with variable length strings for designing a nonparametric classification methodology. Its significance in partitioning different landcover regions from satellite images, having complex/overlapping class boundaries, is demonstrated. The classifier is able to evolve automatically the appropriate number of hyperplanes efficiently for modeling any kind of class boundaries optimally. Merits of the system over the related ones are established through the use of several quantitative measure.
93 citations
TL;DR: Tests using ground data from the Departement Loir-et-Cher, France, have shown that use of the knowledge-based system with GIS data gives an accuracy improvement of approximately 13 per cent compared to using a parametric image classifier alone.
Abstract: This paper describes a knowledge-based system which has been developed for integrating easily-available geographical context information from a GIS in remotely-sensed image analysis. An experiment is described in which soil maps and buffered road networks have been used as additional data layers for classifying single date SPOT images for estimates of crop acreages. The map datasets have been digitised, co-registered to the satellite imagery, and manipulated using ARC/INFO. The knowledge base consists of both image context rules and geographical context rules. Probabilistic information from the image classifier and from the rule base is combined using the Dempster-Shafer model of evidential reasoning. Tests using ground data from the Departement Loir-et-Cher, France, have shown that use of the knowledge-based system with GIS data gives an accuracy improvement of approximately 13 per cent compared to using a parametric image classifier alone.
82 citations