Connectionist Systems Research Articles

FinTech forecasting using an evolving connectionist system for lenders and borrowers: ecosystem behavior

<span lang="EN-US">Financial Technology (FinTech) which is included in the development of digitalization in the financial sector in the industrial era 4.0. Fintech can make any transactions anywhere with the pillars of Peet-to-Peer (P2P) lending, merchants and crowdfunding. In the P2P Lending pillar, there are borrowers and lenders who are digitized in Fintech devices. Fintech in Indonesia is controlled by a state agency called the Otoritas Jasa Keuangan or Financial Services Authority (OJK). In the movement of P2P Lending, there are borrowers and lenders who can be said to be investors where these activities are reported to the OJK. This data can be forecasted using a neural network approach such as ECoS, which is a method capable of forecasting with learning that develops in the hidden layer. In this research article, we present results on forecasting borrowers with a Mean Absolute Percentage Error (MAPE) of 0.148% and forecasting lenders with an accuracy measurement with MAPE of 0.209% with a learning rate 1 = 0.6 and a learning rate 2 = 0.3. So, this forecasting model can be said as an optimization in FinTech activities on the behavior of borrowers and lenders.</span>

Artificial glial cells in artificial neuronal networks: a systematic review

The concept of tripartite synapses has revolutionized the world of neuroscience and the way we understand how information is transmitted in the brain. Since its discovery, some research groups have incorporated into connectionist systems classically focused on the development of Artificial Neuron Networks (ANNs) as a single element, artificial astrocytes that try to optimize performance in problem solving.In this systematic review, we searched the ISI Web of Science for papers that focused on the development of such novel models and their comparison with classical ANNs. A total of 22 papers that satisfied the inclusion criteria were analyzed, showing three different ways of applying the neuromodulatory influence of artificial astrocytes on neural networks. Using Multilayer Perceptron Networks, Artificial Neuro-Glial Newtworks and Multilayer Perceptron with Self-Organizing Maps approaches, a detailed analysis of the incorporation of artificial astrocytic networks has been carried out, and the main differences between the different methods have been weighed up. Regardless of the type of inclusion performed, the greater the complexity of the problem to be solved, it has been observed that the influence of artificial astrocytes has improved the performance of classical ANNs, as occurs in the biological brain.

Tool Degradation Prediction Based on Semimartingale Approximation of Linear Fractional Alpha-Stable Motion and Multi-Feature Fusion

Tool wear will reduce workpieces’ quality and accuracy. In this paper, the vibration signals of the milling process were analyzed, and it was found that historical fluctuations still have an impact on the existing state. First of all, the linear fractional alpha-stable motion (LFSM) was investigated, along with a differential iterative model with it as the noise term is constructed according to the fractional-order Ito formula; the general solution of this model is derived by semimartingale approximation. After that, for the chaotic features of the vibration signal, the time-frequency domain characteristics were extracted using principal component analysis (PCA), and the relationship between the variation of the generalized Hurst exponent and tool wear was established using multifractal detrended fluctuation analysis (MDFA). Then, the maximum prediction length was obtained by the maximum Lyapunov exponent (MLE), which allows for analysis of the vibration signal. Finally, tool condition diagnosis was carried out by the evolving connectionist system (ECoS). The results show that the LFSM iterative model with semimartingale approximation combined with PCA and MDFA are effective for the prediction of vibration trends and tool condition. Further, the monitoring of tool condition using ECoS is also effective.

Reliability-driven time series data analysis in multiple-level deep Learning methods utilizing soft computing methods

This paper introduces a novel method for mining data & information retrieval using series data for the duration. A multi-resolution S transformis viewed as a stage-adjusted transform of wavelet/a parameter window low period. Fourier-transform is used to recover significant characteristics from nonstationary duration series data with electricity-network disruptions. In pattern classification of disturbance waveform information, and incorporated Learning Vector Quantization neural-network & different feed-forward neural-network designs were employed afterwardextending required characteristics from the period of series-data.A fuzzy Multilayer perceptron accepts other connectionist systems and therefore is utilized in the final phase of encoding information inside the linking weights to produce fuzzy disturbance rules pattern inference. With energy signal time series data, a pattern classification performance of 99% was attained. Utilizing the new measurement processes, the data-driven information retrieval was displayed. A method proposed in this research is generically used to mine for information similarities in either duration of the series data pattern.

Evolving Connectionist System and Hidden Semi-Markov Model for Learning-Based Tool Wear Monitoring and Remaining Useful Life Prediction

Tool wear can cause dimensional accuracy and poor surface quality in milling process. During the operation of tool wear, it can also cause breakage and damage of the workpieces. To prevent these conditions, it’s important that the tool wear is monitored and the remaining useful life (RUL) is predicted in real time. In this paper, time domain and frequency domain statistical features are firstly extracted using multi-sensory fusion method, including the cutting force, vibration and acoustic emission sensor. Seven eigenvectors are selected as the input of the prediction model based on the distance correlation coefficient between 140 feature vectors and the wear value, which provide the most sensitive features to wear faults. The paper establishes a nonlinear relationship between high-dimensional feature vectors and tools wear based on the evolving connectionist system (ECoS), which uses the incremental learning algorithm to realize real-time prediction of the tools wear. Finally, using the wear value predicted by ECoS as hidden state sequence of Hidden Semi-Markov Model (HSMM), the RUL prediction of the tool based on HMM is established. The 2010 PHM challenge data were used to train the model. The experimental result shows that in comparison with artificial neural network, the ECoS model has higher prediction accuracy, and its mean RMSE error for three tools is 14.8. In comparison with the RUL prediction of HMM model, Probability-based RUL prediction of HSMM is more stable.

Classification of skin cancer images by applying simple evolving connectionist system

<span id="docs-internal-guid-eea5616b-7fff-5d26-eeb4-1d8c084ec93d"><span>Simple evolving connectionist system (SECoS) is one of data mining classification techniques that recognizing data based on the tested and the training data binding. Data recognition is achieved by aligning testing data to trained data pattern. SECoS uses a feedforward neural network but its hidden layer evolves so that each input layer does not perform epoch. SECoS distance has been modified with the normalized Euclidean distance formula to reduce error in training. This paper recognizes skin cancer by classifying benign malignant skin moles images using SECoS based on parameter combinations. The skin cancer classification has learning rate 1 of 0.3, learning rate 2 of 0.3, sensitivity threshold of 0.5, error threshold of 0.1 and MAPE is 0.5184845 with developing hidden node of 23. Skin cancer recognition by applying modified SECoS algorithm is proven more acceptable. Compared to other methods, SECoS is more robust to error variations.</span></span>

Optimization of Applied Detection Rate in the Simple Evolving Connectionist System Method for Classification of Images Containing Protein

Digital image processing in general to makes images that appear converted to a function of light intensity represented in a two-dimensional plane. The function is a value that will be processed for classification so that the computer is able to recognize the image. Besides classification requires training and testing to produce a small error value and optimal algorithm. The problem of optimization is closely related to the principles and findings of science. Getting the smallest error value by calculating using MAPE for that MAPE calculation is done by using the Detection Rate formula to generalize knowledge in order to find the optimal model. Thus, the application of ANN is very suitable for optimizing classification using the Simple Evolving Connectionist System Method and as the result, the classification of images containing protein with test data is that the eggs work with optimal proof of achieving MAPE without modification of 0.1947% and MAPE which has been modified with the formula detection rate of 0.05554633%.

The QSAR similarity principle in the deep learning era: Confirmation or revision?

Structure–activity relationship (SAR) and quantitative SAR (QSAR) are modeling methods largely used in assessing biological properties of chemical substances. QSAR is based on the hypothesis that the chemical structure is responsible for the activity; it follows that similar molecules are expected to have similar properties. Similarity plays an important role in read across, which categorizes molecules primarily on the basis of similarity. Similarity, and chemical similarity too, is a property differently perceived by humans. The various proposed metrics often disagree with human judgment, and no a unique metric for chemical similarity is universally adopted. Researchers argued that categorization is not only explained by similarity but depends as well on abstract knowledge and the task to accomplish. Moreover, similarity cannot be the unique explanation of a categorization, as different perceptual processes take place in category formation. Assuming that similarity judgments are deeply rooted in human knowledge and perception, cognitive sciences contributions are as important as the mathematical considerations of the classical theories. After an excursus in the many views of similarity in philosophy, mathematics, and cognitive science, the paper explores how connectionist systems, which loosely mimic the human cognitive system, could improve similarity-based choices. A case study on building (Q)SARs using connectionism and deep neural networks shows the role of similarity in building and explaining those models. A discussion about deep learning for QSAR and as a modeling tool for science concludes the presentation.

Isolated word language identification system with hybrid features from a deep belief network

SummaryThe representation of good audio features is the first and foremost requirement for improving the identification performance of any system. Most of the representation learning approaches are based on connectionist systems to learn and extract latent features from the speech data. This research work presents a hybrid feature extraction approach to integrate Mel‐Frequency Cepstral Coefficients (MFCC) features with Shifted Delta Cepstral (SDC) coefficients features, which are further stacked to Deep Belief Network (DBN), for yielding new feature representations of the speech signals. DBN is utilized for unsupervised feature learning on the extracted MFCC‐SDC acoustic features. A 3‐layer Back Propagation Neural Network (BPNN) classifier is initialized in terms of the learning outcomes of hidden layers of DBN for identifying language from the uttered speech. The efficiency of the proposed approach is evaluated by simulating several experimental algorithms on the user‐defined database of isolated words in four languages, namely, Tamil, Malayalam, Hindi, and English, in the working platform of MATLAB. The obtained results for the proposed hybrid approach MFCC‐SDC‐DBN are promising. The proposed approach is also compared with the baseline feature extraction approach MFCC‐SDC by utilizing traditional acoustic features and BPNN classifier. The accuracy obtained with our proposed approach is 98.1% whereas that of the baseline approach is 82%, thereby providing an overall improvement of 16.1%.

Improved Hybrid Swarm Intelligence Algorithm for Voice Command Recognition

Swarm intelligence involves the aggregation of the boids, which interact with each other in their own environment. The boids are agents that follow simple rules in the absence of centralized structure. The Artificial Neural Network is also known as a connectionist system that originated from the biological neural network. The aim of the present study is to improve the hybrid swarm intelligence algorithm for voice command recognition. The proposed algorithm hybrid combines the lion optimization algorithm with the fish swarm algorithm, and was improved upon using voice features. Approximately ten voice commands were recorded, including “open”, “close”, “open door”, “close door”, “open window”, “close window”, “on”, “off”, “play”, and “stop”.

Regression Model and Neural Network Applied to the Public Spending Execution

Artificial Neural Networks are connectionist systems formed by numerous process units called neurons connected to each other, which adapt their structure through learning techniques to solve problems of function approximation and pattern classification. They process information that is supplied to them, either to obtain relationships between them and the objective function that is intended to be approximated, or by classifying these data into different categories. Regression analysis aims to determine the type of functional relationship that exists between a dependent variable and one or more independent variables. The purpose of the research is to use regression methods (multiple regression) and artificial neural networks (multilayer perceptron) to determine the influence of spending execution on the regional government's public budget. 95% of the variability of the budget of Moquegua region has been determined and explained by the three sectors (primary, secondary and tertiary) and 5% is determined by other factors outside the regional government budget. The determination coefficients R2 = 95.9% in the regression model and R2 = 95.3% in the neural network (multilayer perceptron). It has been demonstrated that Artificial neural networks and regression models have obtained very similar results, achieving good and good-fit models.

Evolving connectionist systems (ECoSs): a new approach for modeling daily reference evapotranspiration (ET0)

Over the last few years, the uses of artificial intelligence techniques (AI) for modeling daily reference evapotranspiration (ET0) have become more popular and a considerable amount of models were successfully applied to the problem. Therefore, in the present paper, we propose a new evolving connectionist (ECoS) approaches for modeling daily reference evapotranspiration (ET0) in the Mediterranean region of Algeria. Three ECoS models, namely, (i) the off-line dynamic evolving neural-fuzzy inference system called DEFNIS_OF, (ii) the on-line dynamic evolving neural-fuzzy inference system called DEFNIS_ON, and (iii) the evolving fuzzy neural network called (EFuNN), were statistically compared using the root mean square error (RMSE), the mean absolute error (MAE), the coefficient of correlation (R), and the Nash-Sutcliffe efficiency (NSE) indexes. The proposed approaches were applied for modeling daily ET0 using climatic variables from two weather stations: Algiers and Skikda, Algeria. Five well-known climatic variables were selected as inputs: daily maximum and minimum air temperatures (Tmax and Tmin), daily wind speed (WS), daily relative humidity (RH), and daily sunshine hours (SH). The effect of combining several climatic variables as inputs was evaluated, and at least six scenarios were developed and compared. The proposed ECoS models were compared against the reference Penman-Monteith model referred as "FAO-56 PM". According to the results obtained, the DEFNIS_OF1 model having Tmax, Tmin, WS, RH, and SH as inputs, is the best model, followed by the DEFNIS_ON1, and the EFuNN1 is the worst model. The R and NSE value calculated for the testing dataset for the Algiers and Skikda stations were (0.954, 0.910) and (0.954, 0.905), respectively. While both DEFNIS_OF1 and DEFNIS_ON1 showed good accuracy and high performances, the EFuNN1 was less accurate.

D2 shape distribution and artificial neural networks for 3D objects recognition

In this paper, we propose a 3D object recognition approach, based on the shape distribution D2 and artificial neural networks. The challenge is to discriminate between similar and dissimilar shapes by finding a shape signature that can be constructed and classified quickly. We propose a connectionist system to recognize 3D objects in VRML (Virtual Reality Modeling Language) format. The key idea is to represent the signature of an object as a shape distribution sampled from a shape function measuring global geometric properties of an object. The proposed strategy is the following: from a polygon object to be recognized, a triangulation is performed. Then, distances are calculated between two random points of the triangulated surface of the 3D object. The frequency of these distances will be represented by a normalized histogram. The values of these histograms feed a multi-layer neural network with back- propagation training. We demonstrate the potential of this approach in a set of experiments, which proved our system could achieve above 91.7% recognition rate. In addition, to evaluate the efficiency of our method, we compare our classifier with Support vector machine and k- nearest neighbours. The simulation results highlight the performance of the proposed approach.

Phonological Perspective in Recognizing Kurdish Language Word

This research is confers Speech Perception and a Phonological perspective in a frame of TRACE model in a Connectionist point of view and that’s all to Analysis Kurdish and Persian language relatives. According to this view the research entitled (Speech perception and phonological Analysis in Word Recognition). In this research we assumes that Kurdish and Persian language have many similarity especially in phonological level , for reasoning this assumption and analyzing our data we use TRACE model which, developed by McClelland and Elman (1986), is speech recognition model similar to the interactive model of word recognition , and it is based on a connectionist approach in psycholinguistics.
 The data that this research investigate drive us to serial Conclusion which are close to our first assumption (more relative between two language ) and every difference between two related languages should be explicable to a high degree of plausibility, and systematic changes by this the study declares historical relationships between the two languages and the synchronically difference. The research content a preface and a body which is divide to three sections and a conclusion: First section: it contents a review of Trace Model (concept and the principle) and the reason to work with this model and selecting it among all another models, because TRACE is based on a connectionist system. There are connections among units at three levels: features, phonemes and words and The TRACE model are thus consistent with the idea of competition among units in the lexicon. In other words, at given time.
 Second section: in this section we achieve the TRACE model in Kurdish language, particularly in word cognition and phonological level because the models base on three basic levels: phonological Features, phonemes and word. Third section : we compare in this section both of the two languages(Kurdish and Persian ) through what we got from the results in the second section and this comparative lead us to three main results, like that the both languages have similar phonemic system with a rare different in phonemes feature. 

Measuring the Accuracy of Simple Evolving Connectionist System with Varying Distance Formulas

Simple Evolving Connectionist System (SECoS) is a minimal implementation of Evolving Connectionist Systems (ECoS) in artificial neural networks. The three-layer network architecture of the SECoS could be built based on the given input. In this study, the activation value for the SECoS learning process, which is commonly calculated using normalized Hamming distance, is also calculated using normalized Manhattan distance and normalized Euclidean distance in order to compare the smallest error value and best learning rate obtained. The accuracy of measurement resulted by the three distance formulas are calculated using mean absolute percentage error. In the training phase with several parameters, such as sensitivity threshold, error threshold, first learning rate, and second learning rate, it was found that normalized Euclidean distance is more accurate than both normalized Hamming distance and normalized Manhattan distance. In the case of beta fibrinogen gene -455 G/A polymorphism patients used as training data, the highest mean absolute percentage error value is obtained with normalized Manhattan distance compared to normalized Euclidean distance and normalized Hamming distance. However, the differences are very small that it can be concluded that the three distance formulas used in SECoS do not have a significant effect on the accuracy of the training results.

Semantics of deductive databases with spiking neural P systems

The integration of symbolic reasoning systems based on logic and connectionist systems based on the functioning of living neurons is a vivid research area in computer science. In the literature, one can find many efforts where different reasoning systems based on different logics are linked to classic artificial neural networks. In this paper, we study the relation between the semantics of reasoning systems based on propositional logic and the connectionist model in the framework of membrane computing, namely, spiking neural P systems. We prove that the fixed point semantics of deductive databases without negation can be implemented in the spiking neural P systems model and such a model can also deal with negation if it is endowed with anti-spikes and annihilation rules.

An evolving connectionist system for data stream fuzzy clustering and its online learning

An evolving cascade neuro-fuzzy system and its online learning procedure are considered in this paper. The system is based on conventional Kohonen neurons. The proposed system solves a clustering task of non-stationary data streams under uncertainty conditions when data come in the form of a sequential stream in an online mode. A quality estimation process is defined by finding an optimal value of the used cluster validity index.

FAST MEDICAL DIAGNOSTICS USING AUTOASSOCIATIVE NEURO-FUZZY MEMORY

This paper proposes an architecture of fast medical diagnostics system based on autoassociative neuro-fuzzy memory. The architecture of proposed system is close to traditional Takagi-Sugeno-Kang neuro-fuzzy system, but it is based on other principles. This system contains of recording subsystem and pattern retrieval subsystem, where diagnostics of patients with unknown diagnoses is realized. Level of memberships for all other possible diagnoses from recording subsystem is determined too. System tuning is based on lazy learning procedure and “neurons in data points” principle and uses bell-shaped fuzzy basis functions. Number of these functions changes during training process using principles of evolving connectionist systems. Bell-shaped membership functions centers can be tuned using proposed algorithm, processes of accumulation patients in fundamental memory and patients retrieval are described. This hybrid neuro-fuzzy associative memory combines advantages of fuzzy inference systems, artificial neural networks and evolving systems and its using provides the increasing of autoassociative memories capacity without essential complication of its architecture for medical diagnostics tasks.

A Novel Computer Assisted Automatic Sleep Scoring System by Employing Artificial Neural Network–A review

Sleep is an essential element for an individual’s well-being and is considered vital for the overall mental and physical heath of a person. Sleep can be considered as a virtual detachment of an individual from his environment. In normal humans, about 30% of their life-time is spent for sleep. Artificial neural networks (ANNs) or connectionist systems are computing systems inspired by the biological neural networks that constitute animal brains. Sleep scoring is under taken by the examination and visual inspection of polysomnograms (PSG) done by sleep specialist. PSG is specialty test, the conduction of which includes the recording of various physiological signals. The signals obtained are processed using digital processing tools so as to extract information. Soft computing techniques are used to analyze the signals. ANNs are parallel adaptive systems suitable for solving of non-linear problems. Using ANN for automatic sleep scoring is especially promising because of new ANN learning algorithms allowing faster classification without decreasing the performance. Both appropriate preparations of training data as well as selection of the ANN model make it possible to perform effective and correct recognizing of relevant sleep stages. Such an approach is highly topical, taking into consideration the fact that there is no automatic scorer utilizing ANN technology available at present. The high performances observed with systems based onneural networks highlight that these tools may be act new tools in the field of sleep research. In this scenario we are surmised the review regarding the computer assisted automatic scoring of sleep and soft computing technique Artificial Neural Network.

Speaker recognition with hybrid features from a deep belief network

Learning representation from audio data has shown advantages over the handcrafted features such as mel-frequency cepstral coefficients (MFCCs) in many audio applications. In most of the representation learning approaches, the connectionist systems have been used to learn and extract latent features from the fixed length data. In this paper, we propose an approach to combine the learned features and the MFCC features for speaker recognition task, which can be applied to audio scripts of different lengths. In particular, we study the use of features from different levels of deep belief network for quantizing the audio data into vectors of audio word counts. These vectors represent the audio scripts of different lengths that make them easier to train a classifier. We show in the experiment that the audio word count vectors generated from mixture of DBN features at different layers give better performance than the MFCC features. We also can achieve further improvement by combining the audio word count vector and the MFCC features.