International Journal of Computational Intelligence and Applications

In this paper, the study aims to develop a model for predicting and budgeting maintenance and rehabilitation projects costs for residential buildings throughout their life cycle based on artificial neural network, fuzzy logic and statistical techniques (multi-layer regression). Data consisting of bills of quantities were collected from local Algerian building construction agencies. The principle of costing significant items and work packages has been applied to optimize accurate and efficient document contract model. The results of the research show that neural network has more accuracy with 97% than the multi-layer regression analysis model.

Classification problems in which the number of features (dimensions) is unduly higher than the number of samples (observations) is an essential research and application area in a variety of domains, especially in computational biology. It is also known as a high-dimensional small-sample-size (HDSSS) problem. Various dimensionality reduction methods have been developed, but they are not potent with the small-sample-sized high-dimensional datasets and suffer from overfitting and high-variance gradients. To overcome the pitfalls of sample size and dimensionality, this study employed variational autoencoder (VAE), which is a dynamic framework for unsupervised learning in recent years. The objective of this study is to investigate a reliable classification model for high-dimensional and small-sample-sized datasets with minimal error. Moreover, it evaluated the strength of different architectures of VAE on the HDSSS datasets. In the experiment, six genomic microarray datasets from Kent Ridge Biomedical Dataset Repository were selected, and several choices of dimensions (features) were applied for data preprocessing. Also, to evaluate the classification accuracy and to find a stable and suitable classifier, nine state-of-the-art classifiers that have been successful for classification tasks in high-dimensional data settings were selected. The experimental results demonstrate that the VAE can provide superior performance compared to traditional methods such as PCA, fastICA, FA, NMF, and LDA in terms of accuracy and AUROC.

Selfish herd optimizer (SHO) is a new optimization algorithm. However, its optimization performance is not satisfactory. The main reason for this phenomenon is the weak global search ability of SHO. In this paper, in order to increase the global search ability of SHO, we add Levy-flight distribution strategy. To verify the performance of the proposed algorithm, we use 10 benchmark functions as test cases. Experiment results show that our algorithm is more competitive.

In the context of facial expression recognition (FER), this paper reviews the fundamental theories of emotions and further explains the key dimensions of a defined emotional space. The main contribution of this paper is to propose a set of novel categorization methods for facial expressions to be used in the design of an automatic FER system. This novel categorization enables the facial expression to be interpreted in a better way that and to be more effective in practical applications of automatic FER systems. In order to validate the feasibility of the proposed categorization methods, a set of experiments is reported which investigates and analyzes the influence that the novel categorization brings to a multi-view FER system.

Estimation of software cost and effort is of prime importance in software development process. Accurate and reliable estimation plays a vital role in successful completion of the project. To estimate software cost, various techniques have been used. Constructive Cost Model (COCOMO) is amongst most prominent algorithmic model used for cost estimation. Different versions of COCOMO consider different types of parameters affecting overall cost. Parameters involved in estimation using COCOMO possess vagueness which introduces some degree of uncertainty in algorithmic modelling. The concept of fuzzy logic can deal with uncertainty involved in Intermediate COCOMO cost driver measurements via Fuzzy Inference System (FIS). In the proposed research, an effort has been made wherein, for each cost driver, an FIS is designed to calculate the corresponding effort multiplier. Proposed research provides an insight through evolutionary-based optimization techniques to optimize fuzzy logic-based COCOMO using Particle Swarm Optimization Algorithm. The magnitude of relative error and its mean, calculated using COCOMO NASA2 and COCOMONASA datasets are used as evaluation metrics to validate the proposed model. The model outperforms when compared to other optimization techniques like Genetic Algorithm.

G-Protein-Coupled Receptors (GPCR) are the large family of protein membrane; and until now some of them still remain orphans. Predicting GPCR functions is a challenging task, it depends closely to their classification, which requires a digital representation of each protein chain as an attribute vector. A major problem of GPCR databases is their great number of features which can produce combinatorial explosion and increase the complexity of classification algorithms. Feature selection techniques are used to deal with this problem by minimizing features space dimension, and keeping the most relevant ones. In this paper, we propose to use the BAT algorithm for extracting the pertinent features and to improve the classification results. We compared the results obtained by our system with two other bio-inspired algorithms, Evolutionary Algorithm and PSO search. Metrics quality measures used for comparison are Error Rate, Accuracy, MCC and $F$-measure. Experimental results indicate that our system is more efficient.

The exponential upward push in breast cancer cases across the globe has alarmed academia-industries to obtain certain more effect and strong Breast cancer laptop Aided prognosis (BC-CAD) device for breast most cancers detection. Some of techniques have been evolved with focus on case centric segmentation, feature extraction and class of breast cancer Histopathological photos. However, rising complexity and accuracy regularly demands more sturdy answer. Recently, Convolutional Neural community (CNN) has emerged as one of the maximum efferent techniques for medical records evaluation and diverse picture classification issues. On this paper, a notably strong and green BC-CAD solution has been proposed. Our proposed gadget consists of pre-processing, more suitable adaptive learning based totally Gaussian aggregate model (GMM), connected element analysis based vicinity of interest localization, and AlexNet-DNN primarily based characteristic extraction. The precept factor analysis (PCA) and Linear Discriminant analysis (LDA) primarily based on characteristic selection that's used as dimensional discount. One of the blessings of the proposed method is that not one of the current dimensional reduction algorithms hired with SVM to perform breast most cancers detection and class. The overall results acquired signify that the AlexNet-DNN based capabilities at completely connected layer; FC6 together with LDA dimensional discount and SVM-based totally classification outperforms other country-of-artwork techniques for breast cancer detection. The proposed method completed 96.20 for AlexNet-FC6 and 96.70 for AlexNet-FC7 in term of assessment measures.

This work proposed a memetic version of Artificial Bee Colony algorithm, or called LSABC, which employed a “shrinking” local search strategy. By gradually shrinking the local search space along with the optimization process, the proposed LSABC algorithm randomly explores a large space in the early run time. This helps to avoid premature convergence. Then in the later evolution process, the LSABC finely exploits a small region around the current best solution to achieve a more accurate output value. The optimization behavior of the LSABC algorithm was studied and analyzed in the work. Compared with the classic ABC and several other state-of-the-art optimization algorithms, the LSABC shows a better performance in terms of convergence rate and quality of results for high-dimensional problems.

In this paper, a Sugeno type fuzzy system based on the fuzzy clustering has been developed for a variety of datasets. The number of rules for each dataset is based on the optimum number of clusters in that dataset. Rule sets provide the knowledge base for the classification of data. Each rule set is fine-tuned using the GWO with the intention to improve the classification. The approach is compared with the work of previous researchers on similar data sets using a variety of techniques, including nature-inspired algorithms such as genetic algorithms and Swarm based algorithms. Statistical Analysis of the performance of GWO shows that it is better than five other algorithms 95% of the time.