Computational Intelligence Models Research Articles

Application of computational intelligence methods for complex two-phase flow pattern recognition

Recognizing the significance of accurate predictions of flow conditions in open-channel bends is key in attempts to preserve riverbanks from erosion. The focus of the present work is on the application of two computational intelligence (CI) models to predict free surface flow pattern in a strongly curved 60° open-channel bend. An experimental study is also carried out to prepare the required input and output data set for the CI training and validation process. A set of 476 data is used to train and test the proposed models. The adaptive neuro-fuzzy inference system (ANFIS) and multilayer perceptron (MLP) networks are adopted to construct the models. The CI results are compared with the experimental data obtained. Excellent agreement is found representing the reliability of the employed models. The findings confirm that the models present accurate predictions of the flow depths and velocities, while the MLP network outperforms ANFIS with a mean relative error percentage of 0.67%. New equations are also proposed for estimating the super-elevation and flow of the study.

A novel and reliable computational intelligence system for breast cancer detection.

Cancer is the second important morbidity and mortality factor among women and the most incident type is breast cancer. This paper suggests a hybrid computational intelligence model based on unsupervised and supervised learning techniques, i.e., self-organizing map (SOM) and complex-valued neural network (CVNN), for reliable detection of breast cancer. The dataset used in this paper consists of 822 patients with five features (patient's breast mass shape, margin, density, patient's age, and Breast Imaging Reporting and Data System assessment). The proposed model was used for the first time and can be categorized in two stages. In the first stage, considering the input features, SOM technique was used to cluster the patients with the most similarity. Then, in the second stage, for each cluster, the patient's features were applied to complex-valued neural network and dealt with to classify breast cancer severity (benign or malign). The obtained results corresponding to each patient were compared to the medical diagnosis results using receiver operating characteristic analyses and confusion matrix. In the testing phase, health and disease detection ratios were 94 and 95%, respectively. Accordingly, the superiority of the proposed model was proved and can be used for reliable and robust detection of breast cancer.

Computational intelligence applied to discriminate bee pollen quality and botanical origin

The aim of this work was to develop computational intelligence models based on neural networks (NN), fuzzy models (FM), and support vector machines (SVM) to predict physicochemical composition of bee pollen mixture given their botanical origin. To obtain the predominant plant genus of pollen (was the output variable), based on physicochemical composition (were the input variables of the predictive model), prediction models were learned from data. For the inverse case study, input/output variables were swapped.The probabilistic NN prediction model obtained 98.4% of correct classification of the predominant plant genus of pollen. To obtain the secondary and tertiary plant genus of pollen, the results present a lower accuracy.To predict the physicochemical characteristic of a mixture of bee pollen, given their botanical origin, fuzzy models proven the best results with small prediction errors, and variability lower than 10%.

Using Computational Intelligence Models for Additional Insight into Protein Structure

Proteins are large, complex molecules with crucial roles in the functioning of living organisms. Understanding the underlying mechanisms by which proteins achieve their structures and substructures, as well as those involved in the conformational transitions may contribute to a deeper comprehension of the involved biological processes. This paper investigates a new machine learning perspective upon analyzing protein conformational transitions and introduces a new formalization for the problem, with the more general goal of uncovering interesting patterns in protein conformational transitions. This study represents the starting point of a research which is being conducted in order to obtain a better comprehension of proteins’ structures and, implicitly, functions, by investigating computational intelligence methods for analyzing and deducing proteins conformational transitions.

Computational Intelligence-based PM2.5 Air Pollution Forecasting

Computational intelligence based forecasting approaches proved to be more efficient in real time air pollution forecasting systems than the deterministic ones that are currently applied. Our research main goal is to identify the computational intelligence model that is more proper to real time PM2.5 air pollutant forecasting in urban areas. Starting from the study presented in [27]a, in this paper we first perform a comparative study between the most accurate computational intelligence models that were used for particulate matter (fraction PM2.5) air pollution forecasting: artificial neural networks (ANNs) and adaptive neuro-fuzzy inference system (ANFIS). Based on the obtained experimental results, we make a comprehensive analysis of best ANN architecture identification. The experiments were realized on datasets from the AirBase databases with PM2.5 concentration hourly measurements. The statistical parameters that were computed are mean absolute error, root mean square error, index of agreement and correlation coefficient.

Abstract Intelligence

Basic studies in denotational mathematics and mathematical engineering have led to the theory of abstract intelligence (aI), which is a set of mathematical models of natural and computational intelligence in cognitive informatics (CI) and cognitive computing (CC). Abstract intelligence triggers the recent breakthroughs in cognitive systems such as cognitive computers, cognitive robots, cognitive neural networks, and cognitive learning. This paper reports a set of position statements presented in the plenary panel (Part II) of IEEE ICCI*CC'16 on Cognitive Informatics and Cognitive Computing at Stanford University. The summary is contributed by invited panelists who are part of the world's renowned scholars in the transdisciplinary field of CI and CC.

Computational intelligence models to predict porosity of tablets using minimum features.

The effects of different formulations and manufacturing process conditions on the physical properties of a solid dosage form are of importance to the pharmaceutical industry. It is vital to have in-depth understanding of the material properties and governing parameters of its processes in response to different formulations. Understanding the mentioned aspects will allow tighter control of the process, leading to implementation of quality-by-design (QbD) practices. Computational intelligence (CI) offers an opportunity to create empirical models that can be used to describe the system and predict future outcomes in silico. CI models can help explore the behavior of input parameters, unlocking deeper understanding of the system. This research endeavor presents CI models to predict the porosity of tablets created by roll-compacted binary mixtures, which were milled and compacted under systematically varying conditions. CI models were created using tree-based methods, artificial neural networks (ANNs), and symbolic regression trained on an experimental data set and screened using root-mean-square error (RMSE) scores. The experimental data were composed of proportion of microcrystalline cellulose (MCC) (in percentage), granule size fraction (in micrometers), and die compaction force (in kilonewtons) as inputs and porosity as an output. The resulting models show impressive generalization ability, with ANNs (normalized root-mean-square error [NRMSE] =1%) and symbolic regression (NRMSE =4%) as the best-performing methods, also exhibiting reliable predictive behavior when presented with a challenging external validation data set (best achieved symbolic regression: NRMSE =3%). Symbolic regression demonstrates the transition from the black box modeling paradigm to more transparent predictive models. Predictive performance and feature selection behavior of CI models hints at the most important variables within this factor space.

Active control for traffic lights in regions and corridors: an approach based on evolutionary computation

The growth of vehicles’ fleet circulating on urban streets constitutes a very strong tendency in recent years. The main consequence of this phenomenon refers to the increase of urban congestions, of average delays caused by vehicles waiting on traffic lights and of number of stops. Finding strategies to achieve efficient active traffic control in urban centers is a challenge for engineers and analysts. Recently, important research on dynamic networks and Intelligent Transportation Systems using computational intelligence modeling techniques has been done. This paper proposes a new scheme of active control, using optimization algorithms, to dynamically find traffic signal control plans that optimize traffic conditions in delimited networks and corridors. The proposed system includes a time delay predictive model, used in conjunction with evolutionary approaches like genetic algorithms and differential evolution techniques. Conceptual and applied computational representations necessary for the construction of models are presented. Data collected from a big city in Brazil were fed into the commercial microscopic simulator AIMSUN and were used for the practical experiments. Two main experiments were undertaken and statistically compared in order to decide which method is more efficient in optimizing the active traffic signal timing control for the region under study.

Control of corporate ownership in the evolutional portfolio intelligent complex optimization (EPICO) model

We introduce a new methodology that incorporates advanced higher moments evaluation in a new approach of the Portfolio Selection problem, supported by effective Computational Intelligence models. The Evolutional Portfolio Intelligent Complex Optimization (EPICO) model extracts hidden patterns out of the numerous accounting data and financial statements filtering misguiding effects such as noise or fraud, offering an optimal portfolio selection method.

Day-ahead natural gas demand forecasting based on the combination of wavelet transform and ANFIS/genetic algorithm/neural network model

Accurate forecasts of natural gas demand can be essential for utilities, energy traders, regulatory authorities, decision makers and others. The aim of this paper is to test the robustness of a novel hybrid computational intelligence model in day-ahead natural gas demand predictions. The proposed model combines the Wavelet Transform (WT), Genetic Algorithm (GA), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Feed-Forward Neural Network (FFNN). The WT is used to decompose the original signal in a set of subseries and then a GA optimized ANFIS is employed to provide the forecast for each subseries. ANFIS output is fed into a FFNN to refine the initial forecast and upgrade the overall forecasting accuracy. The model is applied to all distribution points that compose the natural gas grid of a country, in contradiction to the majority of the literature that focuses on a limited number of distribution points. This approach enables the comparison of the model performance on different consumption patterns, providing also insights on the characteristics of large urban centers, small towns, industrial areas, power generation units, public transport filling stations and others.

Intelligence and Its Many Facets an Exploration with a Missing Link

Intelligence is the capacity as well as the ability to learn, develop & apply the acquired knowledge for the development, adjustment & prediction of sequential patterns. Intelligence is the basic ability for measurement of different kinds of intelligence tests. Though positive feelings encourage self-esteem and a sense of worth, conveying that they are important factors for ‘‘intelligent’’ beings. As in1 two principal classifications of intelligence are the operational and the ‘‘real’’ definition. Operational intelligence is measurable with specific aspects. But the real intelligence is one that characterizes the true nature of the thing being defined. Prediction and behaviour are closely related2. But behaviour has emotional impact over it, although intelligence has no direct control on emotions. Hence there must be a link between emotion and intelligence. In3, the appearance of emotional intelligence (EI) into masses served as a missing link in peculiar findings that people with average Intelligence Quotient (IQ) surpass those with the highest IQ. This throws dislodge to the assumption that the sole source of success is only IQ. But emotion is very lubricious as well as habitat dependent. By measuring Emotional Quotient (EQ) of a person in different situations randomly for same purpose and then trying to convert it into wisdom, we could achieve stability. In this manner with the help of wisdom we could proceed from conscious level to advanced consciousness level. A small act of security concern, depicting an individual’s security quotient (SQ), along with emotional content in any social circumstance/domain could be merged into an intelligence activity. The main objective of this paper is to develop an interlink between emotional intelligence quotient (EIQ) & wisdom quotient (WQ) within any specific range of security with recurrent feedback analyzed by self-actualization. If this hypothetical relation is imposed on any computational intelligence model and then experimented with human subjects belonging to various classes, cultures and ages, then our prediction or visualization could be fine-tuned or smoothened to become a general working model.

A Treatise to Computational Approaches Towards Prediction of Membrane Protein and Its Subtypes.

Membrane proteins are vital mediating molecules responsible for the interaction of a cell with its surroundings. These proteins are involved in different functionalities such as ferrying of molecules and nutrients across membrane, recognizing foreign bodies, receiving outside signals and translating them into the cell. Membrane proteins play significant role in drug interaction as nearly 50% of the drug targets are membrane proteins. Due to the momentous role of membrane protein in cell activity, computational models able to predict membrane protein with accurate measures bears indispensable importance. The conventional experimental methods used for annotating membrane proteins are time-consuming and costly and in some cases impossible. Computationally intelligent techniques have emerged to be as a useful resource in the automation of prediction and hence the annotation process. In this study, various techniques have been reviewed that are based on different computational intelligence models used for prediction process. These techniques were formulated by different researchers and were further evaluated to provide a comparative analysis. Analysis shows that the usage of support vector machine-based prediction techniques bears more assiduous results.

Audience real-time bio-signal-processing-based computational intelligence model for narrative scene editing

Narrative scene editing is carried out by directors as an eidetic technique. Continuity editing is a style of editing used in film making to make films as realistic as possible for the audience. While non-continuity editing (e.g., flashbacks, jump cuts, montages, etc.) is also a critical factor that reflects the character of a director, most scenes demand continuity editing to maximize the audience’s narrative immersion. In this paper, we present an algorithm for continuity editing that determines the size of a shot (field of view) by evaluating the psychical distance between the characters and viewers via the measurement of electrodermal activity (or the galvanic skin response). The use of this continuity editing algorithm is expected to result in more audience-friendly videos by reflecting the level of identification between the actors and the audience.

Multiobjective Markov chains optimization problem with strong Pareto frontier: Principles of decision making

In this paper, we present a novel approach for computing the Pareto frontier in Multi-Objective Markov Chains Problems (MOMCPs) that integrates a regularized penalty method for poly-linear functions. In addition, we present a method that make the Pareto frontier more useful as decision support system: it selects the ideal multi-objective option given certain bounds. We restrict our problem to a class of finite, ergodic and controllable Markov chains. The regularized penalty approach is based on the Tikhonov’s regularization method and it employs a projection-gradient approach to find the strong Pareto policies along the Pareto frontier. Different from previous regularized methods, where the regularizator parameter needs to be large enough and modify (some times significantly) the initial functional, our approach balanced the value of the functional using a penalization term (μ) and the regularizator parameter (δ) at the same time improving the computation of the strong Pareto policies. The idea is to optimize the parameters μ and δ such that the functional conserves the original shape. We set the initial value and then decrease it until each policy approximate to the strong Pareto policy. In this sense, we define exactly how the parameters μ and δ tend to zero and we prove the convergence of the gradient regularized penalty algorithm. On the other hand, our policy-gradient multi-objective algorithms exploit a gradient-based approach so that the corresponding image in the objective space gets a Pareto frontier of just strong Pareto policies. We experimentally validate the method presenting a numerical example of a real alternative solution of the vehicle routing planning problem to increase security in transportation of cash and valuables. The decision-making process explored in this work correspond to the most frequent computational intelligent models applied in practice within the Artificial Intelligence research area.

Predicting the Poaceae pollen season: six month-ahead forecasting and identification of relevant features.

In this paper, we approach the problem of predicting the concentrations of Poaceae pollen which define the main pollination season in the city of Madrid. A classification-based approach, based on a computational intelligence model (random forests), is applied to forecast the dates in which risk concentration levels are to be observed. Unlike previous works, the proposal extends the range of forecasting horizons up to 6 months ahead. Furthermore, the proposed model allows to determine the most influential factors for each horizon, making no assumptions about the significance of the weather features. The performace of the proposed model proves it as a successful tool for allergy patients in preventing and minimizing the exposure to risky pollen concentrations and for researchers to gain a deeper insight on the factors driving the pollination season.

Computational intelligence modeling of granule size distribution for oscillating milling

Oscillating mills such as OscilloWitt™ (Frewitt) have been widely used in the secondary manufacture of solid dosage forms in the pharmaceutical industry. This type of mill is generally used for moderate milling of difficult-to-process and heat sensitive materials to a particle size range of c.a. 250μm. Particle size distribution is the result of interaction between ribbon properties and process conditions, therefore it is crucial to model and optimize such a complex process in order to produce more uniform particle size distributions. In this work, multiple linear regression (MLR), genetic programming (GP), and artificial Neural Networks (ANN) assisted by 3-fold cross-validation (CV) were used to present generalized models for the prediction of granule size based on the experimental data set. The normalized mean squared error (NRMSE) and the coefficient of determination (R2) for best fit, namely ANN model were obtained as follows: NRMSE=2.28%, R2=0.9926. MLR model was imprecise in the prediction of d10 class. Due to its performance similarities to ANN and its transparency and ease of application, the GP model could be used widely for granule size prediction. Based on the results it was confirmed that the screen size has the most significant effect on the granule size distribution.

Comparison of computational intelligence models for cuttings transport in horizontal and deviated wells

Improper cleaning of the horizontal and deviated wellbores is one of the major problems encountered during drilling process. Cuttings bed thickness is a key parameter to determine hole cleaning. In this study, computational intelligence techniques including k-Nearest Neighbor (kNN), Support Vector Regression (SVR), Linear Regression and Artificial Neural Network (ANN) were employed to estimate cuttings bed thickness inside a wellbore with and without drill pipe rotation. Cuttings transport measurements for hole inclinations from horizontal to 60°, rates of penetration from 0.0013 to 0.0038m/s, flow velocities from 0.64m/s to 1.59m/s, and pipe rotations from 0 to 120rpm were recorded to train the computational intelligence techniques. More than three hundred different datasets were performed and corresponding cuttings bed thickness values were obtained. The prediction results of computational intelligence techniques were satisfactory, when compared with experimental data. A t-test was conducted and insignificant difference between experimental and predicted data was concluded. ANN provided slightly better performance than the other models when they were compared with their RMSE, AAE, and AAPE values. However, K-nearest neighbor (kNN) can be modeled a lot faster than the other computational intellegence techniques.

Storage time prediction of pork by Computational Intelligence

In this paper, a storage time prediction of pork using Computational Intelligence (CI) model was reported. We investigated a solution based on traditional pork assessment towards a low time-cost parameters acquisition and high accurate CI models by selection of appropriate parameters. The models investigated were built by J48, Naïve Bayes (NB), k-NN, Random Forest (RF), SVM, MLP and Fuzzy approaches. CI input were traditional quality parameters, including pH, water holding capacity (WHC), color and lipid oxidation extracted from 250 samples of 0, 7 and 14days of post mortem. Five parameters (pH, WHC, L∗, a∗ and b∗) were found superior results to determine the storage time and corroborate with identification in minutes. Results showed RF (94.41%), 3-NN (93.57%), Fuzzy Chi (93.23%), Fuzzy W (92.35%), MLP (88.35%), J48 (83.64%), SVM (82.03%) and NB (78.26%) were modeled by the five parameters. One important observation is about the ease of 0-day identification, followed by 14-day and 7-day independently of CI approach. Result of this paper offers the potential of CI for implementation in real scenarios, inclusive for fraud detection and pork quality assessment based on a non-destructive, fast, accurate analysis of the storage time.

Computational Intelligence Modeling of the Macromolecules Release from PLGA Microspheres-Focus on Feature Selection.

Poly-lactide-co-glycolide (PLGA) is a copolymer of lactic and glycolic acid. Drug release from PLGA microspheres depends not only on polymer properties but also on drug type, particle size, morphology of microspheres, release conditions, etc. Selecting a subset of relevant properties for PLGA is a challenging machine learning task as there are over three hundred features to consider. In this work, we formulate the selection of critical attributes for PLGA as a multiobjective optimization problem with the aim of minimizing the error of predicting the dissolution profile while reducing the number of attributes selected. Four bio-inspired optimization algorithms: antlion optimization, binary version of antlion optimization, grey wolf optimization, and social spider optimization are used to select the optimal feature set for predicting the dissolution profile of PLGA. Besides these, LASSO algorithm is also used for comparisons. Selection of crucial variables is performed under the assumption that both predictability and model simplicity are of equal importance to the final result. During the feature selection process, a set of input variables is employed to find minimum generalization error across different predictive models and their settings/architectures. The methodology is evaluated using predictive modeling for which various tools are chosen, such as Cubist, random forests, artificial neural networks (monotonic MLP, deep learning MLP), multivariate adaptive regression splines, classification and regression tree, and hybrid systems of fuzzy logic and evolutionary computations (fugeR). The experimental results are compared with the results reported by Szlȩk. We obtain a normalized root mean square error (NRMSE) of 15.97% versus 15.4%, and the number of selected input features is smaller, nine versus eleven.

Prediction of Pathological Stage in Patients with Prostate Cancer: A Neuro-Fuzzy Model.

The prediction of cancer staging in prostate cancer is a process for estimating the likelihood that the cancer has spread before treatment is given to the patient. Although important for determining the most suitable treatment and optimal management strategy for patients, staging continues to present significant challenges to clinicians. Clinical test results such as the pre-treatment Prostate-Specific Antigen (PSA) level, the biopsy most common tumor pattern (Primary Gleason pattern) and the second most common tumor pattern (Secondary Gleason pattern) in tissue biopsies, and the clinical T stage can be used by clinicians to predict the pathological stage of cancer. However, not every patient will return abnormal results in all tests. This significantly influences the capacity to effectively predict the stage of prostate cancer. Herein we have developed a neuro-fuzzy computational intelligence model for classifying and predicting the likelihood of a patient having Organ-Confined Disease (OCD) or Extra-Prostatic Disease (ED) using a prostate cancer patient dataset obtained from The Cancer Genome Atlas (TCGA) Research Network. The system input consisted of the following variables: Primary and Secondary Gleason biopsy patterns, PSA levels, age at diagnosis, and clinical T stage. The performance of the neuro-fuzzy system was compared to other computational intelligence based approaches, namely the Artificial Neural Network, Fuzzy C-Means, Support Vector Machine, the Naive Bayes classifiers, and also the AJCC pTNM Staging Nomogram which is commonly used by clinicians. A comparison of the optimal Receiver Operating Characteristic (ROC) points that were identified using these approaches, revealed that the neuro-fuzzy system, at its optimal point, returns the largest Area Under the ROC Curve (AUC), with a low number of false positives (FPR = 0.274, TPR = 0.789, AUC = 0.812). The proposed approach is also an improvement over the AJCC pTNM Staging Nomogram (FPR = 0.032, TPR = 0.197, AUC = 0.582).