• A systematic literature review of machine learning applications for community-acquired pneumonia

      Lozano-Rojas, Daniel; Free, Robert C.; McEwan, Alistair A.; Woltmann, Gerrit; University of Leicester; University of Derby; University Hospitals of Leicester NHS Trust, Leicester (Springer, 2021-08-15)
      Community acquired pneumonia (CAP) is an acute respiratory disease with a high mortality rate. CAP management follows clinical and radiological diagnosis, severity evaluation and standardised treatment protocols. Although established in practice, protocols are labour intensive, time-critical and can be error prone, as their effectiveness depends on clinical expertise. Thus, an approach for capturing clinical expertise in a more analytical way is desirable both in terms of cost, expediency, and patient outcome. This paper presents a systematic literature review of Machine Learning (ML) applied to CAP. A search of three scholarly international databases revealed 23 relevant peer reviewed studies, that were categorised and evaluated relative to clinical output. Results show interest in the application of ML to CAP, particularly in image processing for diagnosis, and an opportunity for further investigation in the application of ML; both for patient outcome prediction and treatment allocation. We conclude our review by identifying potential areas for future research in applying ML to improve CAP management. This research was co-funded by the NIHR Leicester Biomedical Research Centre and the University of Leicester.
    • Criminal networks analysis in missing data scenarios through graph distances

      Ficara, Annamaria; Cavallaro, Lucia; Curreri, Francesco; Fiumara, Giacomo; De Meo, Pasquale; Bagdasar, Ovidiu; Song, Wei; Liotta, Antonio; University of Palermo, Palermo, Italy; University of Messina, Messina, Italy (Public Library of Science (PLoS), 2021-08-11)
      Data collected in criminal investigations may suffer from issues like: (i) incompleteness, due to the covert nature of criminal organizations; (ii) incorrectness, caused by either unintentional data collection errors or intentional deception by criminals; (iii) inconsistency, when the same information is collected into law enforcement databases multiple times, or in different formats. In this paper we analyze nine real criminal networks of different nature (i.e., Mafia networks, criminal street gangs and terrorist organizations) in order to quantify the impact of incomplete data, and to determine which network type is most affected by it. The networks are firstly pruned using two specific methods: (i) random edge removal, simulating the scenario in which the Law Enforcement Agencies fail to intercept some calls, or to spot sporadic meetings among suspects; (ii) node removal, modeling the situation in which some suspects cannot be intercepted or investigated. Finally we compute spectral distances (i.e., Adjacency, Laplacian and normalized Laplacian Spectral Distances) and matrix distances (i.e., Root Euclidean Distance) between the complete and pruned networks, which we compare using statistical analysis. Our investigation identifies two main features: first, the overall understanding of the criminal networks remains high even with incomplete data on criminal interactions (i.e., when 10% of edges are removed); second, removing even a small fraction of suspects not investigated (i.e., 2% of nodes are removed) may lead to significant misinterpretation of the overall network.
    • A collaborative approach for national cybersecurity incident management

      Oriola, Oluwafemi; Adeyemo, Adesesan Barnabas; Papadaki, Maria; Kotzé, Eduan; university of Plymouth; University of Ibadan, Ibadan, Nigeria; University of the Free State, Bloemfontein, South Africa (Emerald, 2021-06-28)
      Collaborative-based national cybersecurity incident management benefits from the huge size of incident information, large-scale information security devices and aggregation of security skills. However, no existing collaborative approach has been able to cater for multiple regulators, divergent incident views and incident reputation trust issues that national cybersecurity incident management presents. This paper aims to propose a collaborative approach to handle these issues cost-effectively. A collaborative-based national cybersecurity incident management architecture based on ITU-T X.1056 security incident management framework is proposed. It is composed of the cooperative regulatory unit with cooperative and third-party management strategies and an execution unit, with incident handling and response strategies. Novel collaborative incident prioritization and mitigation planning models that are fit for incident handling in national cybersecurity incident management are proposed. Use case depicting how the collaborative-based national cybersecurity incident management would function within a typical information and communication technology ecosystem is illustrated. The proposed collaborative approach is evaluated based on the performances of an experimental cyber-incident management system against two multistage attack scenarios. The results show that the proposed approach is more reliable compared to the existing ones based on descriptive statistics. The approach produces better incident impact scores and rankings than standard tools. The approach reduces the total response costs by 8.33% and false positive rate by 97.20% for the first attack scenario, while it reduces the total response costs by 26.67% and false positive rate by 78.83% for the second attack scenario.
    • Internet of Planets (IoP): A New Era of the Internet

      Kang, Byungseok; Malute, Francis; Bagdasar, Ovidiu; Hong, Choongseon; University of Derby; Kyung Hee University, Seoul, South Korea (Institute of Electrical and Electronics Engineers (IEEE), 2021-06-24)
      Internet of Planets (IoP) is a concept that enables solar planets to communicate with each other using the Internet. While there is a plethora of research on IoP, the delay tolerant network (DTN) has emerged as the most advanced technology in recent years. DTN is an asynchronous networking technology that has been deployed for the networking environment in which steady communication paths are not available, and therefore, it stores receiving data in a data storage and forward them only when the communication links are established. DTN can be applied to sensor networks and the mobile ad-hoc network, as well as space communication that supports data transmissions among satellites. In DTN networking environments, it is crucial to secure a scheme that has relatively low routing overhead and high reliability to achieve efficiency. Thus, this article proposes a time (delay) information based DTN routing scheme, which is able to predict routing paths for achieving efficient data transmissions among the nodes that have comparatively periodic moving patterns. The results of the proposed DTN routing algorithm using NS-3 simulation tools indicate satisfied levels of routing performance in comparison with the existing DTN algorithm.
    • Research on Action Strategies and Simulations of DRL and MCTS-based Intelligent Round Game

      Sun, Yuxiang; Yuan, Bo; Zhang, Yongliang; Zheng, Wanwen; Xia, Qingfeng; Tang, Bojian; Zhou, Xianzhong; Nanjing University, China; University of Derby; Army Engineering University, Nanjing, China (Springer Science and Business Media LLC, 2021-06-16)
      The reinforcement learning problem of complex action control in multiplayer online battlefield games has brought considerable interest in the deep learning field. This problem involves more complex states and action spaces than traditional confrontation games, making it difficult to search for any strategy with human-level performance. This paper presents a deep reinforcement learning model to solve this problem from the perspective of game simulations and algorithm implementation. A reverse reinforcement-learning model based on high-level player training data is established to support downstream algorithms. With less training data, the proposed model is converged quicker, and more consistent with the action strategies of high-level players’ decision-making. Then an intelligent deduction algorithm based on DDQN is developed to achieve a better generalization ability under the guidance of a given reward function. At the game simulation level, this paper constructs Monte Carlo Tree Search Intelligent Decision Model for turn-based antagonistic deduction games to generate next-step actions. Furthermore, a prototype game simulator that combines offline with online functions is implemented to verify the performance of proposed model and algorithm. The experiments show that our proposed approach not only has a better reference value to the antagonistic environment using incomplete information, but also accurate and effective in predicting the return value. Moreover, our work provides a theoretical validation platform and testbed for related research on game AI for deductive games.
    • Recommender Systems Evaluator: A Framework for Evaluating the Performance of Recommender Systems

      dos Santos, Paulo V.G.; Tardiole Kuehne, Bruno; Batista, Bruno G.; Leite, Dionisio M.; Peixoto, Maycon L.M.; Moreira, Edmilson Marmo; Reiff-Marganiec, Stephan; University of Derby; Federal University of Itajubá, Itajubá, Brazil; Federal University of Mato Grosso do Sul (UFMS), Ponta Porã, Brazil; et al. (Springer, 2021-06-05)
      Recommender systems are filters that suggest products of interest to customers, which may positively impact sales. Nowadays, there is a multitude of algorithms for recommender systems, and their performance varies widely. So it is crucial to choose the most suitable option given a situation, but it is not a trivial task. In this context, we propose the Recommender Systems Evaluator (RSE): a framework aimed to accomplish an offline performance evaluation of recommender systems. We argue that the usage of a proper methodology is crucial when evaluating the available options. However, it is frequently overlooked, leading to inconsistent results. To help appraisers draw reliable conclusions, RSE is based on statistical concepts and displays results intuitively. A comparative study of classical recommendation algorithms is presented as an evaluation, highlighting RSE’s critical features.
    • An empirical analysis of the information security culture key factors framework

      Tolah, Alaa; Furnell, Steven; Papadaki, Maria; University of Plymouth; Saudi Electronic University, Riyadh, Saudi Arabia; University of Nottingham; University of Derby; Nelson Mandela University, Gqeberha, South Africa (Elsevier, 2021-06-05)
      Information security is a challenge facing organisations, as security breaches pose a serious threat to sensitive information. Organisations face security risks in relation to their information assets, which may also stem from their own employees. Organisations need to focus on employee behaviour to limit security failures, as if they wish to establish effective security culture with employees acting as a natural safeguard for information assets. This study was conducted to respond to a need for more empirical studies that focus on a development of security culture to provide a comprehensive framework. The Information Security Culture and Key Factors Framework has been developed, incorporating two types of factors: those that influence security culture and those that reflect it. This paper validates the applicability of the framework and tests related hypotheses through an empirical study. An exploratory survey was conducted, and 266 valid responses were obtained. Phase two of the study demonstrates the framework levels of validity and reliability through the use of factor analysis. Different hypothetical correlations were analysed through the use of structural equation modelling, with indirect exploratory effect of the moderators achieved through a multi-group analysis. The findings show that the framework has validity and achieved an acceptable fit with the data. This study fills an important gap in the significant relationship between personality traits and security culture. It also contributes to the improvement of information security management through the introduction of a comprehensive framework in practice, which functions in the establishment of security culture. The factors are vital in justifying security culture acceptance, and the framework provides an important tool that can be used to assess and improve an organisational security culture.
    • COVID-19 pandemic decision support system for a population defense strategy and vaccination effectiveness

      Varotsos, Costas A; Krapivin, Vladimir F; Xue, Yong; Soldatov, Vladimir; Voronova, Tatiana; National and Kapodistrian University of Athens, Athens, Greece; Kotelnikov’s Institute of Radioengineering and Electronics, Fryazino Branch, Russian Academy of Sciences, Vvedensky 1, Fryazino, Moscow Region 141190, Russian Federation; University of Mining and Technology, Xuzhou, Jiangsu 221116, PR China; University of Derby (Elsevier BV, 2021-06-05)
      The year 2020 ended with a significant COVID-19 pandemic, which traumatized almost many countries where the lockdowns were restored, and numerous emotional social protests erupted. According to the World Health Organization, the global epidemiological situation in the first months of 2021 deteriorated. In this paper, the decision-making supporting system (DMSS) is proposed to be an epidemiological prediction tool. COVID-19 trends in several countries and regions, take into account the big data clouds for important geophysical and socio-ecological characteristics and the expected potentials of the medical service, including vaccination and restrictions on population migration both within the country and international traffic. These parameters for numerical simulations are estimated from officially delivered data that allows the verification of theoretical results. The numerical simulations of the transition and the results of COVID-19 are mainly based on the deterministic approach and the algorithm for processing statistical data based on the instability indicator. DMSS has been shown to help predict the effects of COVID-19 depending on the protection strategies against COVID-19 including vaccination. Numerical simulations have shown that DMSS provides results using accompanying information in the appropriate scenario.
    • Nowcasting of air pollution episodes in megacities: A case study for Athens, Greece

      Varotsos, Costas A.; Mazei, Yuri; Saldaev, Damir; Efstathiou, Maria; Voronova, Tatiana; Xue, Yong; University of Athens, Athens, Greece; Lomonosov Moscow State University, Leninskiye Gory, 1, Moscow, Russia; Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia; Shenzhen MSU-BIT University, Shenzhen, China; et al. (Elsevier BV, 2021-06-02)
      The main objective of the present study is to develop a model for the prediction of the extreme events of air pollution in megacities using the concept of so-called "natural time" instead of the "conventional clock time". In particular, we develop a new nowcasting technique based on a statistically significant fit to the law of Gutenberg-Richter of the surface concentration of ozone (O3), particles of the size fraction less than 10 μm (PM-10) and nitrogen dioxide (NO2). Studying the air pollution over Athens, Greece during the period 2000–2018, we found that the average waiting time between successive extreme concentrations values varied between different atmospheric parameters accounted as 17 days in case of O3, 29 days in case of PM-10 and 28 days in case of NO2. This average waiting time depends on the upper threshold of the maximum extreme concentrations of air pollutants considered. For instance, considering the NO2 concentrations over Athens it was found that the average waiting time is 13 days for 130 μg/m3 and 2.4 years for 200 μg/m3. Remarkably, the same behaviour of obedience to the Guttenberg-Richter law characterizing the extreme values of the air pollution of a megacity was found earlier in other long-term ecological and paleoclimatic variables. It is a sign of self-similarity that is often observed in nature, which can be used in the development of more reliable nowcasting models of extreme events.
    • Electro-Thermal Coupled Modeling of Induction Motor Using 2D Finite Element Method

      Bousbaine, Amar; Bouheraoua, Mustapha; Atig, M.; Benamrouche, N; University of Derby; Université Mouloud Mammeri de Tizi Ouzou (Ştefan cel Mare University of Suceava, 2021-05-31)
      The paper evaluates the thermal behavior of an induction machine based on a coupled electromagnetic-thermal model using 2D non-linear complex finite element method. The currents and the temperature distribution in a squirrel cage induction motor in transient state are investigated and presented. The convection heat transfer coefficient between the frame and ambient and the windings are treated with particular attention. The developed method can be applied to other electric machines having negligible axial heat flow. The corroboration of the theoretical/simulated results have been investigated, experimentally using a 2.2 kW totally enclosed fan-cooled induction motor. The simulated results and those obtained from measurements have been critically evaluated and showed good agreements.
    • Performance evaluation of machine learning techniques for fault diagnosis in vehicle fleet tracking modules

      Sepulevene, Luis; Drummond, Isabela; Kuehne, Bruno Tardiole; Frinhani, Rafael; Filho, Dionisio Leite; Peixoto, Maycon; Reiff-Marganiec, Stephan; Batista, Bruno; Federal University of Itajubá, Itajubá, Brazil; Federal University of Mato Grosso do Sul, Ponta Porã, Brazil; et al. (Oxford University Press, 2021-05-14)
      With industry 4.0, data-based approaches are in vogue. However, extracting the essential features is not a trivial task and greatly influences the fi nal result. There is also a need for specialized system knowledge to monitor the environment and diagnose faults. In this context, the diagnosis of faults is signi cant, for example, in a vehicle fleet monitoring system, since it is possible to diagnose faults even before the customer is aware of the fault, minimizing the maintenance costs of the modules. In this paper, several models using Machine Learning (ML) techniques were applied and analyzed during the fault diagnosis process in vehicle fleet tracking modules. Two approaches were proposed, "With Knowledge" and "Without Knowledge", to explore the dataset using ML techniques to generate classi fiers that can assist in the fault diagnosis process. The approach "With Knowledge" performs the feature extraction manually, using the ML techniques: Random Forest, Naive Bayes, Support Vector Machine (SVM) and Multi Layer Perceptron (MLP); on the other hand, the approach "Without Knowledge" performs an automatic feature extraction, through a Convolutional Neural Network (CNN). The results showed that the proposed approaches are promising. The best models with manual feature extraction obtained a precision of 99.76% and 99.68% for detection and detection and isolation of faults, respectively, in the provided dataset. The best models performing an automatic feature extraction obtained respectively 88.43% and 54.98% for detection and detection and isolation of failures.
    • On k-partitions of multisets with equal sums

      Andrica, Dorin; Bagdasar, Ovidiu; Babeş-Bolyai University of Cluj-Napoca, Cluj-Napoca, Romania; University of Derby (Springer Science and Business Media LLC, 2021-05-05)
      We study the number of ordered k-partitions of a multiset with equal sums, having elements α1,…,αn and multiplicities m1,…,mn. Denoting this number by Sk(α1,…,αn;m1,…,mn), we find the generating function, derive an integral formula, and illustrate the results by numerical examples. The special case involving the set {1,…,n} presents particular interest and leads to the new integer sequences Sk(n), Qk(n), and Rk(n), for which we provide explicit formulae and combinatorial interpretations. Conjectures in connection to some superelliptic Diophantine equations and an asymptotic formula are also discussed. The results extend previous work concerning 2- and 3-partitions of multisets.
    • On Generalized Lucas Pseudoprimality of Level k

      Andrica, Dorin; Bagdasar, Ovidiu; Babeş-Bolyai University, 400084 Cluj-Napoca, Romania; University of Derby (MDPI AG, 2021-04-12)
      We investigate the Fibonacci pseudoprimes of level k, and we disprove a statement concerning the relationship between the sets of different levels, and also discuss a counterpart of this result for the Lucas pseudoprimes of level k. We then use some recent arithmetic properties of the generalized Lucas, and generalized Pell–Lucas sequences, to define some new types of pseudoprimes of levels k+ and k− and parameter a. For these novel pseudoprime sequences we investigate some basic properties and calculate numerous associated integer sequences which we have added to the Online Encyclopedia of Integer Sequences.
    • Targeted ensemble machine classification approach for supporting IOT enabled skin disease detection

      Yu, Hong Qing; Reiff-Marganiec, Stephan; University of Derby (IEEE, 2021-03-26)
      The fast development of the Internet of Things (IoT) changes our life in many areas, especially in the health domain. For example, remote disease diagnosis can be achieved more efficiently with advanced IoT-technologies which not only include hardware but also smart IoT data processing and learning algorithms, e.g. image-based disease classification. In this paper, we work in a specific area of skin condition classification. This research work aims to provide an implementable solution for IoT-led remote skin disease diagnosis applications. The research output can be concluded into three folders. The first folder is about dynamic AI model configuration supported IoT-Fog-Cloud remote diagnosis architecture with hardware examples. The second folder is the evaluation survey regarding the performances of machine learning models for skin disease detection. The evaluation contains a variety of data processing methods and their aggregations. The evaluation takes account of both training-testing and cross-testing validations on all seven conditions and individual condition. In addition, the HAM10000 dataset is picked for the evaluation process according to the suitability comparisons to other relevant datasets. In the evaluation, we discuss the earlier work of ANN, SVM and KNN models, but the evaluation process mainly focuses on six widely applied Deep Learning models of VGG16, Inception, Xception, MobileNet, ResNet50 and DenseNet161. The result shows that each of the top four models for the major seven skin conditions has better performance for the specific condition than others. Based on the evaluation discovery, the last folder proposes a novel classification approach of the Targeted Ensemble Machine Classify Model (TEMCM) to enable dynamically combining a suitable model in a two-phase detection process. The final evaluation result shows the proposed model can archive better performance.
    • Pseudoprimality related to the generalized Lucas sequences

      Andrica, Dorin; Bagdasar, Ovidiu; Babeş-Bolyai University, Cluj-Napoca, Romania; University of Derby (Elsevier BV, 2021-03-13)
      Some arithmetic properties and new pseudoprimality results concerning generalized Lucas sequences are presented. The findings are connected to the classical Fibonacci, Lucas, Pell, and Pell–Lucas pseudoprimality. During the process new integer sequences are found and some conjectures are formulated.
    • An LMI Approach-Based Mathematical Model to Control Aedes aegypti Mosquitoes Population via Biological Control

      Dianavinnarasi, J.; Raja, R.; Alzabut, J.; Niezabitowski, M.; Selvam, G.; Bagdasar, O.; Alagappa University, Karaikudi 630 004, India; Prince Sultan University, Riyadh 12435, Saudi Arabia; Silesian University of Technology, Akademicka 16, 44-100, Gliwice, Poland; Vinayaka Missions University, Salem 636308, India; et al. (Hindawi Limited, 2021-03-09)
      In this paper, a novel age-structured delayed mathematical model to control Aedes aegypti mosquitoes via Wolbachia-infected mosquitoes is introduced. To eliminate the deadly mosquito-borne diseases such as dengue, chikungunya, yellow fever, and Zika virus, the Wolbachia infection is introduced into the wild mosquito population at every stage. This method is one of the promising biological control strategies. To predict the optimal amount of Wolbachia release, the time varying delay is considered. Firstly, the positiveness of the solution and existence of both Wolbachia present and Wolbachia free equilibrium were discussed. Through linearization, construction of suitable Lyapunov–Krasovskii functional, and linear matrix inequality theory (LMI), the exponential stability is also analyzed. Finally, the simulation results are presented for the real-world data collected from the existing literature to show the effectiveness of the proposed model.
    • Controlling Wolbachia transmission and invasion dynamics among aedes aegypti population via impulsive control strategy

      Dianavinnarasi, Joseph; Raja, Ramachandran; Alzabut, Jehad; Niezabitowski, Michał; Bagdasar, Ovidiu; Alagappa University, Karaikudi, India; Prince Sultan University, Riyadh, Saudi Arabia; Silesian University of Technology, Akademicka 16, Gliwice, Poland; University of Derby (MDPI AG, 2021-03-08)
      This work is devoted to analyzing an impulsive control synthesis to maintain the self-sustainability of Wolbachia among Aedes Aegypti mosquitoes. The present paper provides a fractional order Wolbachia invasive model. Through fixed point theory, this work derives the existence and uniqueness results for the proposed model. Also, we performed a global Mittag-Leffler stability analysis via Linear Matrix Inequality theory and Lyapunov theory. As a result of this controller synthesis, the sustainability of Wolbachia is preserved and non-Wolbachia mosquitoes are eradicated. Finally, a numerical simulation is established for the published data to analyze the nature of the proposed Wolbachia invasive model.
    • Graph and Network Theory for the Analysis of Criminal Networks

      Cavallaro, Lucia; Bagdasar, Ovidiu; De Meo, Pasquale; Fumara, Giacomo; Liotta, Antonio; University of Derby; University of Messina, Italy; Free University of Bozen-Bolzano, Italy (Springer, Cham, 2021-02-19)
      Social Network Analysis is the use of Network and Graph Theory to study social phenomena, which was found to be highly relevant in areas like Criminology. This chapter provides an overview of key methods and tools that may be used for the analysis of criminal networks, which are presented in a real-world case study. Starting from available juridical acts, we have extracted data on the interactions among suspects within two Sicilian Mafia clans, obtaining two weighted undirected graphs. Then, we have investigated the roles of these weights on the criminal networks properties, focusing on two key features: weight distribution and shortest path length. We also present an experiment that aims to construct an artificial network which mirrors criminal behaviours. To this end, we have conducted a comparative degree distribution analysis between the real criminal networks, using some of the most popular artificial network models: Watts-Strogats, Erdős-Rényi, and Barabási-Albert, with some topology variations. This chapter will be a valuable tool for researchers who wish to employ social network analysis within their own area of interest.
    • Application of caputo–fabrizio operator to suppress the aedes aegypti mosquitoes via wolbachia: an LMI approach

      Dianavinnarasi, J.; Raja, R.; Alzabut, J.; Cao, J.; Niezabitowski, M.; Bagdasar, O.; Alagappa University, Karaikudi, India; Prince Sultan University, Riyadh 12435, Saudi Arabia; Southeast University, Nanjing, China; Yonsei University, Seoul, South Korea; et al. (Elsevier BV, 2021-02-11)
      The aim of this paper is to establish the stability results based on the approach of Linear Matrix Inequality (LMI) for the addressed mathematical model using Caputo–Fabrizio operator (CF operator). Firstly, we extend some existing results of Caputo fractional derivative in the literature to a new fractional order operator without using singular kernel which was introduced by Caputo and Fabrizio. Secondly, we have created a mathematical model to increase Cytoplasmic Incompatibility (CI) in Aedes Aegypti mosquitoes by releasing Wolbachia infected mosquitoes. By this, we can suppress the population density of A.Aegypti mosquitoes and can control most common mosquito-borne diseases such as Dengue, Zika fever, Chikungunya, Yellow fever and so on. Our main aim in this paper is to examine the behaviours of Caputo–Fabrizio operator over the logistic growth equation of a population system then, prove the existence and uniqueness of the solution for the considered mathematical model using CF operator. Also, we check the alpha-exponential stability results for the system via linear matrix inequality technique. Finally a numerical example is provided to check the behaviour of the CF operator on the population system by incorporating the real world data available in the known literature.
    • Blessing of dimensionality at the edge and geometry of few-shot learning

      Tyukin, Ivan Y.; Gorban, Alexander N.; McEwan, Alistair A.; Meshkinfamfard, Sepehr; Tang, Lixin; University of Leicester; Lobachevsky University, Russia; St Petersburg State Electrotechnical University, Russia; University College London; Northeastern University, China; et al. (Elsevier BV, 2021-02-03)
      In this paper we present theory and algorithms enabling classes of Artificial Intelligence (AI) systems to continuously and incrementally improve with a priori quantifiable guarantees – or more specifically remove classification errors – over time. This is distinct from state-of-the-art machine learning, AI, and software approaches. The theory enables building few-shot AI correction algorithms and provides conditions justifying their successful application. Another feature of this approach is that, in the supervised setting, the computational complexity of training is linear in the number of training samples. At the time of classification, the computational complexity is bounded by few inner product calculations. Moreover, the implementation is shown to be very scalable. This makes it viable for deployment in applications where computational power and memory are limited, such as embedded environments. It enables the possibility for fast on-line optimisation using improved training samples. The approach is based on the concentration of measure effects and stochastic separation theorems and is illustrated with an example on the identification faulty processes in Computer Numerical Control (CNC) milling and with a case study on adaptive removal of false positives in an industrial video surveillance and analytics system.