Pattern Recognition Problems Research Articles

On the Convergence of Tsetlin Machines for the IDENTITY- and NOT Operators.

The Tsetlin Machine (TM) is a recent machine learning algorithm with several distinct properties, such as interpretability, simplicity, and hardware-friendliness. Although numerous empirical evaluations report on its performance, the mathematical analysis of its convergence is still open. In this article, we analyze the convergence of the TM with only one clause involved for classification. More specifically, we examine two basic logical operators, namely, the "IDENTITY"- and "NOT" operators. Our analysis reveals that the TM, with just one clause, can converge correctly to the intended logical operator, learning from training data over an infinite time horizon. Besides, it can capture arbitrarily rare patterns and select the most accurate one when two candidate patterns are incompatible, by configuring a granularity parameter. The analysis of the convergence of the two basic operators lays the foundation for analyzing other logical operators. These analyses altogether, from a mathematical perspective, provide new insights on why TMs have obtained state-of-the-art performance on several pattern recognition problems.

COVID-19 CORONAVIRUS SCREENING ANALYSIS NEURAL NETWORK TECHNOLOGY

Low-cost, reliable and quick screening diagnosis of coronavirus can be implemented on the basis of intelligent technologies for analyzing a set of signs and symptoms with solving the problem of pattern recognition in the basis of artificial neural networks. The high degree of coronavirus infection diagnostic procedure uncertainty, the vector dimension of input factor-symptoms, fuzzy conditioning and poor formalizability of the subject condition connection with these symptoms require appropriate analytical tools. An analysis of the problem and possible solutions allows justifying the feasibilit y of implementing screening diagnostics as a solution to the problem of nonlinear optimization in a multidimensional space of high-dimensional factors and states. Artificial neural networks with compulsory training on a representative sample were chosen as a tool for implementing the project. The proposed technology brings diagnostics of coronavirus infection closer to full automation, robotization and intellectualization of complex monitoring (diagnostic) systems as the most promising technology for pattern recognition in systems with a high degree of entropy and allows you to solve the problem at the lowest cost and required performance indicators.

Artificial intelligence using for medical diagnosis via implementation of expert systems

Modern biomedical technologies development affords to provide the doctor with colossal amount of information about patients organism condition. However, the opportunity of using this data for medical diagnosis fully now is a distantive perspective only. The reason is a humans limited ability in assessment and interpretation this data arrays. The solution seems in artificial intelligence and expert systems wide introduction to medicine. Currently, almost all authors consider various options for constructing artificial neural networks as a way to implement artificial intelligence. This approach, which goes back to the fundamental theorem of A.N. Kolmogorov, the works of V.I. Arnold and Hecht-Nielsen [3], demonstrates excellent capabilities in a number of pattern recognition problems, which are reduced to revealing hidden details against the background of input noises. Much less often is mentioned such a method of modeling formal thinking as expert systems, which arose in the 1960s and then went into the shadows. Since the inception of cybernetics, computer programmers have tried to reproduce the mechanism of human thinking, that is, the task was to teach the computer to "think". The first known results in the field of creating and using intelligent systems were laid by the work of Norbert Wiener and G.S. Altshuller. At the same time, the creation of intelligent systems was reduced to the development of programs that solve problems using a variety of heuristic methods based on the property of human thinking to generalize.

A fast and accurate circle detection algorithm based on random sampling

Circle detection in digital images is an important problem in computer vision, pattern recognition, and artificial intelligence. However, the common circle detection strategies, including random sample consensus, randomized Hough transform, and randomized circle detection, have a very low sampling efficiency and thus a slow detection speed, owing to aimless random sampling. This paper proposes a fast and accurate randomized circle detection algorithm, with the aim to improve the speed and accuracy of circle detection based on random sampling. The proposed algorithm mainly focuses on four aspects: calculating circle parameters, determining candidate circles, searching for true circle, and improving detection accuracy. To verify the effectiveness of our algorithm, contrastive experiments were conducted on lots of synthetic and real images. The results show that our algorithm achieved much higher detection speed and accuracy than random sample consensus, randomized Hough transform, and randomized circle detection, and realized similar robustness as the three contrastive strategies. The research ideas can also be applied to ellipse detection.

Improving the Performance of an Associative Classifier in the Context of Class-Imbalanced Classification

Class imbalance remains an open problem in pattern recognition, machine learning, and related fields. Many of the state-of-the-art classification algorithms tend to classify all unbalanced dataset patterns by assigning them to a majority class, thus failing to correctly classify a minority class. Associative memories are models used for pattern recall; however, they can also be employed for pattern classification. In this paper, a novel method for improving the classification performance of a hybrid associative classifier with translation (better known by its acronym in Spanish, CHAT) is presented. The extreme center points (ECP) method modifies the CHAT algorithm by exploring alternative vectors in a hyperspace for translating the training data, which is an inherent step of the original algorithm. We demonstrate the importance of our proposal by applying it to imbalanced datasets and comparing the performance to well-known classifiers by means of the balanced accuracy. The proposed method not only enhances the performance of the original CHAT algorithm, but it also outperforms state-of-the-art classifiers in four of the twelve analyzed datasets, making it a suitable algorithm for classification in imbalanced class scenarios.

Can One Design a Series of Fast, Energy - Efficient “Brains” based on Neuromorphic Computing to Solve Complex Pattern Recognition Problem?

Background: In this paper, we present a theoretical discussion on neuromorphic computing circuit dynamic and its relations with AI deep learning neural networks. The hardware implememtations of neuromorphic computing and AI deep learning neuronal networks are discussed. Objective: The investigation is motivated by the design of a feasible fast and energy-efficient circuit device as well as an efficient training computation method to solve complex classification problems using AI neuronal networks. Methods: We focus on the investigations of solving pattern classification and recognition problems in real applications from the perspectives of both logic computation view and physical circuit device views. FPGA approaches are considered and a mapping from logic level to physical level is proposed. Results: A pragmatic mapping method is derived. FPGA method is proposed. Conclusion: Thus we propose in this paper an approach to solve complex classification problems. First, the neuromorphic computing as a new research area is introduced, including physical circuit properties, memristive device physical properties and the circuit dynamics described by the temporal and spatial (Maxwell) differential equations. Secondly, we show that by using AI deep learning neural networks to train AI neural networks we are able to derive the optimal AI neuron network weights. Last but not the least, we brief a mapping method and show in general how the neuromorphic circuit will work in practice after mapping the weights from AI deep learning neural networks into the neuromorphic circuit synapses/memristors. We also devote our discussions to the physical device feasibility and related matters. The method proposed in this paper is pragmatic and constructive.

A novel solder joint diagnosis algorithm using differential geometry approach in high-density flexible integrated circuit substrates

To provide a system with high-density solder joint 3D information acquisition and improve the comprehensive performance of automatic optical inspection systems, a novel diagnosis system for solder joints of high-density flexible integrated circuit substrates is proposed in this paper. The differential geometry principle is introduced to analyze the appearance identity of solder joints, and the diagnosis is handled as a pattern-recognition problem with the Frenet frame-matching approach. First, a high-precision 3D information-acquisition system constituted of two CCD digital cameras, each with a microscope and a three-color LED illumination, is developed. Second, a discrete estimation scheme based on the differential geometry theory is proposed to achieve curvature measurement, and then the shape of the solder joints is characterized by the crest line representation method. Finally, the classification strategy is to establish different models of solder joint types and make a decision based on a set of the predefined diagnosis rules. By inspecting various types of solder joints and comparing with other typical algorithms, the performance of the proposed system is evaluated. The experiment results indicate that the proposed algorithm has advantages of fast inspection speed, low misjudgment rate, which is reduced to 2.2%, and detailed classification.

A Comparison Study on Shape Parameter Selection in Pattern Recognition by Radial Basis Function Neural Networks

This study investigates three choices of shape parameter selection when the so-called Radial Basis Function (RBF) is used. Under the problem of pattern recognition via RBF-Neural Network using RC-algorithm, three RBFs are focussed on; Gaussian (GA), Multiquadric (MQ), and Compactly-Supported (CS1). Two pattern recognition cases are tested and the best choice of shape parameter is validated using Model-Selection Criteria (MSC).

Development of Artificial Intelligence and Prospects for Its Application

Aim. To analyze the concept of “artificial intelligence”, to justify the effectiveness of using artificial intelligence technologies.Tasks. To study the conceptual apparatus; to propose and justify the author’s definition of the “artificial intelligence” concept; to describe the technology of speech recognition using artificial intelligence.Methodology. The authors used such general scientific methods of cognition as comparison, deduction and induction, analysis, generalization and systematization.Results. Based on a comparative analysis of the existing conceptual apparatus, it is concluded that there is no single concept of “artificial intelligence”. Each author puts his own vision into it. In this regard, the author’s definition of the “artificial intelligence” concept is formulated. It is determined that an important area of applying artificial intelligence technologies in various fields of activity is speech recognition technology. It is shown that the first commercially successful speech recognition prototypes appeared already by the 1990s, and since the beginning of the 21st century. The great interest in “end-to-end” automatic speech recognition has become obvious. While traditional phonetic approaches have requested pronunciation, acoustic, and language model data, end-to-end models simultaneously consider all components of speech recognition, thereby facilitating the stages of self-learning and development. It is established that a significant increase in the” mental “ capabilities of computer technology and the development of new algorithms have led to new achievements in this direction. These advances are driven by the growing demand for speech recognition.Conclusions. According to the authors, artificial intelligence is a complex of computer programs that duplicate the functions of the human brain, opening up the possibility of informal learning based on big data processing, allowing to solve the problems of pattern recognition (text, image, speech) and the formation of management decisions. Currently, the active development of information and communication technologies and artificial intelligence concepts has led to a wide practical application of intelligent technologies, especially in control systems. The impact of these systems can be found in the work of mobile phones and expert systems, in forecasting and other areas. Among the obstacles to the development of this technology is the lack of accuracy in speech and voice recognition systems in the conditions of sound interference, which is always present in the external environment. However, the recent advances overcome this disadvantage.

Augmentation on CNNs for Handwritten Digit Classification in a Small Training Sample Size Situation

In general, a deep learning needs a lot of samples. However, in a practical pattern recognition problem, the number of training samples is usually limited. We investigate the effect of an image data augmentation by a perspective transformation on a convolution neural network(CNN) for handwritten digit classification in a small training sample size situation. The experimental results show the effectiveness of the image data augmentation by the perspective transformation on the CNN for handwritten digit classification particularly in the small training sample size situation.

Improved algorithm for matched-pairs selection of informative features in the problems of recognition of complex system states

The problem of computer diagnostics of complex systems is one of the non-trivial tasks of modern information technology. Such systems are, for example, computer networks, automatic and/or automated control systems for complex technological objects, including related to complex problems of environmental protection, biology, etc. In pattern recognition, one of the major problems is forming subspaces of informative features, which only in the «ensemble» allow diagnosing the states of such systems with a high degree of reliability. An effective approach to solving this problem based on the principles of inductive modeling of complex systems is proposed. The quality criterion for recognizing classes of patterns is formulated, which also makes it possible to evaluate the quality of the constructed ensemble of informative features. As an example, the problem of constructing an ensemble of informative features represented by a binary code based on the data of an experiment to determine the hazard levels of some plant protection products is considered. Real primary data on plant protection products used in practice were applied to recognize the effect of certain characteristics on the so-called integrated «hazard indicator». Comparative numerical estimates of the effectiveness of the proposed approach are given. In this case, there can be a fivefold gain in the amount of computations for a relatively small number of input features equal to 5 compared to the known algorithms of the class considered in the paper. It is shown that, from a practical point of view, the described algorithm has advantages over the known algorithms with brute-force search of feature subspaces in pattern recognition problems.

A Biomimetic Covering Learning Method Based on Principle of Homology Continuity

The Principle of Homology Continuity (PHC) based covering learning method is an effective method to solve the pattern recognition problem. However, PHC and the existence of optimal coverage are not mathematical proven. To address this issue, we firstly give the mathematical description and theoretical proof of PHC. On this basis, the theoretical definition of optimal coverage is introduced. Optimal coverage can determine the internal connections among samples as prior knowledge and use covering neurons to learn prior knowledge. Finally, we propose a kind of covering neuron model, and the effectiveness of which is demonstrated through extensive experiments conducted on the CIFAR-10, LFW, and YTF datasets.

Adaptive Learning for Robust Radial Basis Function Networks.

This article addresses the robust estimation of the output layer linear parameters in a radial basis function network (RBFN). A prominent method used to estimate the output layer parameters in an RBFN with the predetermined hidden layer parameters is the least-squares estimation, which is the maximum-likelihood (ML) solution in the specific case of the Gaussian noise. We highlight the connection between the ML estimation and minimizing the Kullback-Leibler (KL) divergence between the actual noise distribution and the assumed Gaussian noise. Based on this connection, a method is proposed using a variant of a generalized KL divergence, which is known to be more robust to outliers in the pattern recognition and machine-learning problems. The proposed approach produces a surrogate-likelihood function, which is robust in the sense that it is adaptive to a broader class of noise distributions. Several signal processing experiments are conducted using artificially generated and real-world data. It is shown that in all cases, the proposed adaptive learning algorithm outperforms the standard approaches in terms of mean-squared error (MSE). Using the relative increase in the MSE for different noise conditions, we compare the robustness of our proposed algorithm with the existing methods for robust RBFN training and show that our method results in overall improvement in terms of absolute MSE values and consistency.

Writer identification using redundant writing patterns and dual-factor analysis of variance

Writer identification (WI) is a typical pattern recognition problem with the goal of recognizing the writer of a text from images of his or her handwriting. For handwriting-based applications, a new approach is required that can confirm the writer based on a very small amount of available text. However, due to the limited number of characters and the diverse contents of the available samples, it is still challenging to achieve a high recognition rate on large-scale databases. To solve this problem, this paper proposes an efficient method for offline text-independent WI using redundant writing patterns and dual-factor analysis of variance (DF-ANOVA). In this method, a search for writing patterns in a limited number of handwritten texts is first conducted to generate meaningful codebooks. Then, these writing patterns are described by means of the simplified Wigner distribution function (WDF) and improved directional index histogram (DIH) descriptors to extract both structural and texture features, and the two corresponding feature distances are integrated by means of a fuzzy integral rule to make the final decision. The Fisher discriminant ratio (FDR) is used for feature selection, and DF-ANOVA is used to eliminate the influence of factors associated with the labels of samples on the feature distance. The results of evaluations on the IAM (96.92%), Firemaker (96.4%), and Uyghur2016 (100%) databases illustrate that this system shows strong robustness to an increasing number of writers and a decreasing number of words in the sample texts.

Machine Learning-Assisted Array-Based Detection of Proteins in Serum Using Functionalized MoS2 Nanosheets and Green Fluorescent Protein Conjugates.

Abnormal concentrations of a specific protein or the presence of some biomarker proteins may indicate life-threatening diseases. Pattern-based detection of specific analytes using affinity-regulated receptors is one of the potential alternatives to specific antigen–antibody-based detection. In this report, we have schemed a sensor array by using various functionalized two-dimensional (2D)-MoS2 nanosheets and green fluorescent protein (GFP) as the receptor and the signal transducer, respectively. Two-dimensional MoS2 has been used as a promising candidate for recognition of the bioanalytes because of its high surface-to-volume ratio compared to those of other nanomaterials. Easy surface tunability of this material provides additional advantages to analyze the target of interest. The optimized 2D-MoS2–GFP conjugates are able to discriminate 15 different proteins at 50 nM concentration with a detection limit of 1 nM. Moreover, proteins in the binary mixture and in the presence of serum were discriminated successfully. Ten different proteins in serum media at relevant concentrations were classified successfully with 100% jackknifed classification accuracy, which proves the potentiality of the above system. We have also implemented and discussed the implication of using different machine learning models on the pattern recognition problem associated with array-based sensing.

Analysis of the features of image processing using the Hamming network on the STM-32 microcontroller

The paper discusses the features of solving a class of problems for pattern recognition using the STM-32 microcontroller. The problem of pattern recognition can be solved on neural networks of different architectures, the main attention is paid to the Hamming neural network model. The features of the implementation of the Hamming network based on the STM-32 microcontroller for the recognition of images entered via the touch screen are analyzed. It is experimentally shown that the network cannot always correctly process the input value and compare it with the reference value of the class for digital test images. This is due to the high degree of similarity of some images and the presence of noise. In conclusion, recommendations on the implementation of neural network algorithms for image processing on microcontrollers are given.

Research on Image Rotation Recognition Based on Histogram Feature Description

Image recognition is a statistical pattern recognition problem in pattern recognition, which mainly includes four steps: image acquisition, preprocessing, feature extraction and selection, and pattern classification. In this paper, an omni-directional sequential feature description method based on F histogram is proposed. In this method, shapes are segmented by directional lines in all directions, the sequential feature description quantity between object segmentation segments in each direction line is calculated, the omni-directional sequential feature description quantity of shapes is constructed, and the similarity between shapes is calculated. In this method, the normal vector histogram is used as the point cloud feature descriptor, which can optimize the search of the corresponding point set, thus improving the ICP algorithm. By selecting the appropriate expansion center and expansion order, the recognition ability of the filter is improved. At the same time, using the concept of image moment, the characteristic moment vector of the target is constructed to identify the target with in-plane rotation and scale change. Based on the optimization strategy based on histogram feature description, the recognition speed of the algorithm is significantly improved.

Research of the efficiency of application of technologies of wavelet analysis in the tasks of recognition of images

The work is devoted to the experimental comparison of the accuracy of classification methods on the problem of pattern recognition in images using wavelet analysis technologies and without. In particular, the interaction of the ring-projection wavelet-fractal method for identifying features with classical classification methods such as "Naive Bayes classifier" and "Support vector machines" is investigated. The experimental test results are given in the form of a table. As a result, it is established that the introduction of wavelet analysis into the construction of image classification models is justified, and leads to a relatively small but significant increase in the classification accuracy.

A New Approach to Solving the Problem of Generating Sets of Complex Structural Objects Based on a Quasi-Equivalent Transformation of a Labeling Scheme

The paper presents the results of a theoretical study related to the development of methods for constructing generating structures based on labeling schemes for generating sets of complex structural objects. In a theoretical aspect, generated objects are mappings of sets of objects into a set of labels, and in practical terms, they can be, in particular, visual images. The scientific and practical interest in generative constructions is that they can be used to determine whether objects belong to a certain class, that is, to solve the problem of pattern recognition. The problem of constructing generating labeling scheme belongs to a wide section of modern applied informatics that embraces Constraint Satisfaction Problem and related themes [1–4]. But this problem has not been posed before and there are still no regular methods for solving it. The analysis of the above methods is based on the formalism of the consistent labeling problem [6, 10, 11], which is, on the one hand, a generalization of many statements of discrete problems of Constraint Satisfaction, and, on the other hand, a transparent theoretical construction with a well-developed mathematical foundation. The problem of constructing a relational scheme (in this case, labeling scheme) that generates a given set of mappings, by analogy with linguistic models, may be named “the problem of grammar restoration” [12–14]. In previous studies it was shown that to solve this problem it makes sense to use equivalent transformations of the labeling scheme [11]. This is because the source table listing all the complex objects that should be generated by the target scheme is itself a trivial variant of the scheme with a given set of consistent labelings. This means that the source scheme and target scheme are equivalent. However, one of the equivalent operations – disunion of a column – cannot be used regularly, since it requires certain conditions to be met regarding the internal structure of the column. In this case, to expand the capabilities of four known equivalent transformations of the labeling scheme – deleting and appending nonexistent labeling, as well as joining of columns and column disunion – a non-equivalent transformation was added – "coloring the column labelings". The purpose of the paper is to introduce and investigate operation of "coloring the column labelings" that leads to a non-equivalent transformation of a labeling scheme. Show the advisability of using the known equivalent and the introduced quasi-equivalent transformations of the labeling scheme to solve the problem of constructing generating structures based on labeling schemes. Results. The transformation of the labeling scheme, called "coloring the labelings of the scheme column", has been introduced. It is shown that its implementation leads to a quasi-equivalent labeling scheme, by solving which it is possible to uniquely restore the solution of the original problem. A method is proposed for using the newly introduced operation to transform the labeling scheme into a quasi-equivalent labeling scheme, in which it becomes possible to regularly perform the column decoupling operation. This ability of the operation of "coloring the column labelings" opens the way to the creation of a method for solving the problem of restoring a labeling scheme that generates a given set of consistent labelings. Keywords: relational scheme, consistent labeling scheme, equivalent labeling scheme transformations, constraint satisfaction problem.

Research on nonlinear forecast and influencing factors of foreign trade export based on support vector neural network

As a kind of time series, the export volume of foreign trade has the characteristics of randomness, complexity, strong nonlinearity and noise, so it is difficult to describe it by using the traditional time series model algorithm. Support vector neural network (SVNN) has many unique advantages in solving small sample, nonlinear and high-dimensional pattern recognition problems. This paper uses the method of support vector neural network to predict and analyze China's foreign trade export and uses principal component analysis and regression analysis to analyze the contribution rate of different influencing factors to foreign trade export. The results show that:(1) the contribution rate of domestic economic factors to China's foreign trade export is the largest, reaching 59.65%, which also reflects the necessity and correctness of China's insistence on supply-side reform. (2) The nonlinear prediction results of the support vector neural network have a good fitting with the actual value of China's foreign trade export, and the prediction error of the support vector neural network is controlled within 10%, showing a good prediction effect; (3) neural network method has good modeling and generalization ability for nonstationary small sample import and export time series data and can achieve high prediction accuracy and decision judgment accuracy, especially for the prediction of its development trend, and the model has a high degree of fitting.