Powerful Machine Research Articles

The legacy of the Maxim machine gun in postcolonial Africa: Implications for governance and management scholarship

ABSTRACT The essay draws on aspects of British colonial history to throw light on Africa’s postcolonial state and society. Under the leadership of Captain Lugard, drawing from the principles of the dual mandate and indirect rule, colonialists employed the then most powerful Maxim machine gun whose legacy of coercive governance set the stage for the rise of the African gunmaster, establishment of gunmaster regime culture, leaving the postcolonial state and society without humanistic socio-cultural value systems such as erade. The essay briefly discusses some of the implications of the legacy for governance and management scholarship and points to new areas and directions for research using multidisciplinary approaches. These include development of a Pan-African humanistic socio-cultural value system, paying more attention to gunmaster regimes and socio-cultural value systems in society, the concept of privilege, and the study of princelings as a unique and possibly developmental cohort.

Single- and Multiswitch Fault-Tolerant Inverter Topology With Preserved Output Power and Extreme Learning Machine Fault Detector

Single- and Multiswitch Fault-Tolerant Inverter Topology With Preserved Output Power and Extreme Learning Machine Fault Detector

Framework for real-time computation and streaming of soundscapes

The auralization of dynamic soundscapes brings significant computational challenges, as multiple often moving sources have to be considered simultaneously. Making such auralizations interactive rules out pre-rendering, requiring the underlying computations to be performed in real-time instead. The “Virtual Acoustics” (VA) framework allows for such real-time applications by leveraging geometrical acoustics and efficient multi-threading. Delivering these simulations to a larger audience, however, has proven difficult, as they still require a lengthy setup and possible simulation complexity is highly dependent on available hardware. When also including high-fidelity 3D visualization, hardware requirements rise further, hindering remote deployment. We present a web-based approach that executes simulations on a server and streams the result in real-time, offloading heavy calculations to a sufficiently powerful machine. 3D graphics are rendered by Unreal Engine and streamed using Pixel Streaming. VA auralization is transmitted via WebRTC audio streams. Both visualization and auralization are ultimately aggregated in a single website for platform-independent access. User input is gathered by the website and relaid back to the server, making the experience interactive. Thereby, this approach can deliver high-fidelity virtual environments with minimal setup and hardware requirements for the user. The approach will be demonstrated by streaming an interactive city park soundscape.

Price-Predicting Model: Predictions for All Sailboats

Sailboat pricing is a nuanced field, significantly influenced by vessel age, geographical location, and brand reputation. To address this complexity, we leveraged the XGBoost algorithm, a powerful machine learning tool, and refined it through Bayesian optimization to create a highly accurate model for predicting sailboat prices. Our analysis revealed significant regional disparities, with sailboats from Europe, the USA, and the Caribbean commanding varying market valuations. When applying this optimized model to the Hong Kong used sailboat market, we were able to generate forecasts with improved accuracy, though still with room for enhancement. Notably, the Lagoon brand stood out as a consistent high performer, underscoring the importance of brand reputation in sailboat pricing. To provide actionable intelligence, we have synthesized a comprehensive analysis that integrates our research findings and predictive strategies. This analysis presents clear visualizations alongside concise interpretations, thereby constituting a pivotal resource for informed decision-making within the local market context.

Development and validation of a machine learning model integrated with the clinical workflow for inpatient discharge date prediction.

Discharge date prediction plays a crucial role in healthcare management, enabling efficient resource allocation and patient care planning. Accurate estimation of the discharge date can optimize hospital operations and facilitate better patient outcomes. In this study, we employed a systematic approach to develop a discharge date prediction model. We collaborated closely with clinical experts to identify relevant data elements that contribute to the prediction accuracy. Feature engineering was used to extract predictive features from both structured and unstructured data sources. XGBoost, a powerful machine learning algorithm, was employed for the prediction task. Furthermore, the developed model was seamlessly integrated into a widely used Electronic Medical Record (EMR) system, ensuring practical usability. The model achieved a performance surpassing baseline estimates by up to 35.68% in the F1-score. Post-deployment, the model demonstrated operational value by aligning with MS GMLOS and contributing to an 18.96% reduction in excess hospital days. Our findings highlight the effectiveness and potential value of the developed discharge date prediction model in clinical practice. By improving the accuracy of discharge date estimations, the model has the potential to enhance healthcare resource management and patient care planning. Additional research endeavors should prioritize the evaluation of the model's long-term applicability across diverse scenarios and the comprehensive analysis of its influence on patient outcomes.

Which classifiers are connected to others? An optimal connection framework for multi-layer ensemble systems

Ensemble learning is a powerful machine learning strategy that combines multiple models e.g. classifiers to improve predictions beyond what any single model can achieve. Until recently, traditional ensemble methods typically use only one layer of models which limits the exploration of different aspects in the classifiers’ predictions. On the other hand, the rise of deep learning has introduced multi-layer architectures that can learn complex functions by transforming data into multiple levels of representation. This characteristic of deep learning suggests that multi-layer ensembles may potentially provide better performance compared to single-layer ensembles. However, a problem which might arise is that in the subsequent layers, not all the inputs to a classifier are desirable, leading to lower performance. In this paper, we introduce a novel multi-layer ensemble of classifiers named COME in which each classifier at a specific layer is connected to multiple classifiers in the previous layer. These connections signify the use of the previous-layer-classifiers’ outputs as inputs for training the current layer's classifier. Each classifier can be connected to different classifiers in the previous layer, which allows inputs in each layer to be optimally selected. We propose a binary encoding scheme to encode the topology of the proposed multi-layer ensemble with defined connections between layers. Differential Evolution, a popular evolutionary computation method, is used as the optimisation algorithm to search for the optimal set of connections. Experimental results on 30 datasets from the UCI Machine Learning Repository and OpenML demonstrate that our proposed ensemble outperforms many state-of-the-art ensemble learning algorithms.

Revolutionizing Logistics and Fleet Management: A Comprehensive Analysis of the Impact of EunoKinetiX on Operational Efficiency and Societal Dynamics

The purpose of this paper is to analyze the impacts of EunoKinetiX, an Enterprise Resource Planning SaaS ( Software as a Service ) in the Fleet Management spheres coupled with Route Optimization for more efficient Logistical provision. EunoKinetiX is a platform intended to assist Logistical providers manage their services and resources, both machine and human power. The product employs artificial intelligence, both predictive and generative for route optimization and payload allocation, effectively reducing costs, CO2 emissions, and saving valuable time. Through advanced analytics, it streamlines logistics, enhancing operational efficiency while contributing to environmental sustainability. The product's integration of AI technology showcases its potential to revolutionize contemporary fleet management practices, offering a compelling solution to disorganized systems to maximize profits.

Open AI and its Impact on Fraud Detection in Financial Industry

As per the Nilson report, fraudulent activities targeting cards amounted to a loss of $32.34 billion globally in 2021, a 14 % increase from the previous year. Such practices can be combated by harnessing OpenAI’s powerful machine learning and automation capabilities. Such advanced technologies help financial companies avoid any potential fraud and protect their esteemed clients' interests. Through the adoption and utilization of such innovative technologies., financial institutions will be better placed to protect their customers and entities from financial losses. Digital fraudsters are skilful in identifying loopholes and have developed cunning techniques like phishing for unsuspecting victims and wittingly swindling money off them. They are also updated in using OpenAI to develop deceitful information to scam people. This has seen the emergence of names like WormGPT and FraudGPT, reliant on generative AI models used by tech corporations with fraud intents. As a result, fraud detection techniques have to evolve with time as fraudsters progressively devise new techniques that bypass old and rigid banking security protocols and learn how to convince unsuspecting individuals to dispatch their money to them.

Prototyping disruptive self-sufficiency power machines composed by cascaded power units based on thermo-hydraulic actuators

This research focuses on developing a prototype for a disruptive Self-Sustaining Power Machine (SSPM) that consists of cascaded power units (PUs). Each PU integrates thermo-mechanical and thermo-hydraulic actuators to drive hydraulic pumps connected to generators. The prototype's design leverages two thermal cycle types, sVsVs and VsVs, which enable self-sustaining operations by employing several strategic capabilities that challenge the limitations of second-kind perpetual motion machines (PMMs). Therefore, key capabilities include: 1- Performing useful mechanical work through the expansion and contraction of the Thermal Working Fluid (TWF) by adding and removing heat. 2- Utilizing recovered heat through potential superposition techniques, thereby upgrading the heat to a higher grade, which enhances heat recovery strategies. 3- Designing an efficient energy transfer chain from thermal to electrical energy via thermo-mechanical or thermo-hydraulic actuators, a hydraulic energy drive system, and conversion from hydraulic to electrical energy. 4- Configuring the system to achieve more useful work than the heat added to the initial PU downstream. Two main prototype types are proposed to demonstrate these proposed capabilities: 1- Continuous motion using the sVsVs thermal cycle, with volumetric reservoirs for thermal working fluids (TWF) located outside the thermo-mechanical or thermo-hydraulic actuators. 2- Discontinuous motion using the VsVs thermal cycle, with volumetric reservoirs for TWF located inside the active volume of the thermo-mechanical or thermo-hydraulic actuators. The prototypes were evaluated through case studies using air and helium as real TWFs. Results indicate that the SSPM composed of five cascaded PUs operating on an sVsVs cycle achieved a global efficiency of 149.6% (SSI of 49.6%) with helium and 126% (SSI of 26%) with air. In contrast, the VsVs cycle yielded a global efficiency of 183% (SSI of 83%) with helium and 133% (SSI of 33%) with air

Financial Fraud Detection and Prevention

Financial fraud has always been a serious challenge in the financial sector. With the continuous development of technology, fraud in the financial market has become increasingly complex and hidden. Therefore, financial fraud detection and prevention have become particularly important to our lives. But as of today, the financial fraud detection methods that have emerged often leave something to be desired. Traditional detection methods based on rules and statistical methods perform poorly when processing large-scale and high-dimensional data, and are prone to false positives and false negatives. Moreover, as fraud techniques continue to evolve, the adaptability of traditional methods is also challenged. As a powerful machine learning technology, deep learning has excellent feature extraction and pattern recognition capabilities, and has achieved remarkable achievements in various fields, including image recognition, natural language processing, and speech recognition. In the financial field, the application of deep learning is also gradually emerging.

Polyconvex inelastic constitutive artificial neural networks

AbstractThe characterization of the material behavior of inelastic materials requires a high degree of expert knowledge to identify and constitutively describe the material response. In addition, specific models are usually pre‐selected in the course of characterization and only the best parameters for these specific models are determined, but therefore not necessarily the best models. Unfortunately, a more general description of these pre‐selected models results in an increased effort during characterization, which is barely practicable by hand. This is where machine learning algorithms may help us. To get the best of both worlds, powerful machine learning and sound thermodynamic considerations, inelastic Constitutive Artificial Neural Networks (iCANNs) discover generic formulations of the Helmholtz free energy and pseudo potential. Constitutive relations guide us towards thermodynamically consistent descriptions of stresses and inelastic strains; a concept that is applicable to a wide range of inelastic phenomena from viscoelasticity, elastoplasticity, phase transformations to growth and remodeling of living tissues. Here, we equip the original iCANN framework to guarantee polyconvexity a priori, which ensures at least one minimizing deformation. We investigate the ability of our iCANN to identify the material behavior of a viscoelastic polymer at different stretch levels and strain rates. We made our source code, data, and example accessible to the public at https://doi.org/10.5281/zenodo.11084354.

Simulated analysis and experimental validation of sound field distribution regularity for a large-scale non-anechoic harbor pool

The large-scale underwater space, such as large water pool or large-scale harbor pool can provide a good experimental environment for the measurement of the underwater radiated noise of vessel. However, it's not as simple as measuring the sound power of a small machine in the air, the measurement and evaluation of the low frequency mechanical noise and radiated sound power of vessel requires a specific reverberation environment in the large-scale underwater space. Therefore, in this paper, the reverberant sound field characteristics of large-scale underwater space are simulated and calculated by finite element analysis, and experimental measurements are carried out in the large-scale harbor pool. It was found the cut-off frequency of large-scale underwater space depends on the depth of the water pool, and the lower limit frequency depends on the volume of the water pool. The dock door does not reduce the spatial average sound pressure in the reverberation sound field, the floating box platform does not effect on the cut-off frequency but has an effect on the sound energy of the reverberant sound field. It is possible to select the appropriate underwater experimental environment for the measurement and evaluation of different mechanical noise sources. The sound pressure level obtained by calculating the sound source at different locations in the reverberation area is basically the same, and in a 16 × 22 hydrophone arrays, more than 160 array elements are selected, each between 0.25 m and 1 m, a stable sound pressure level can be obtained through spatial averaging. Through experimental measurement verification in the large-scale harbor pool, low-frequency sound sources are emitted from 6 different positions, three hydrophone arrays data are collected, and the measured sound pressure level error is within 1.4 dB. These conclusions are conducive to the actual measurement and evaluation of noise sources of vessel.

Hygienic assessment of the protective properties of personal hand protection equipment during vibration testing

Introduction. In industrial environments, at the workplaces of hand tool operators, it is not always possible to completely eliminate increased levels of vibration exposure. The power and speed characteristics of modern hand-held machines are constantly increasing. The impact of local vibration on a person has a negative effect on his body. When working with hand tools, personal protective equipment is used. The purpose of personal hand protection against vibration is to reduce the vibration impact on a person. Materials and methods. Bench tests were carried out on hand protection products consisting of damping materials of various shapes and locations in the product. The vibration sensor was installed and fastened in places where the operator’s hands came into contact with vibration and on the forearm. The effectiveness of protection was determined by the difference in measurement levels. Results. During the tests, vibration levels were recorded on the handle of the vibration stand, the palmar surface, the fingers, and the operator’s forearm. In the low-frequency region of the spectrum, vibration attenuation is minimal, and as the frequency range increases, the attenuation increases. Across all areas of the spectrum, performance levels for mittens are higher than for gloves. In places where the fingers come into contact with vibration, the efficiency is less than when in contact with the hand, and in the low-frequency region of the spectrum a negative efficiency of up to 0.8 dB is noted. Vibration measurements on the operator’s forearm showed vibration levels in the forearm to be lower than in the hand. Limitations. The study was conducted over a period of one year, in laboratory conditions, and was limited to two thousand measurements. Conclusion. Hygienic assessment of the effectiveness of the protective properties of personal protective equipment for hands was carried out in accordance with developed laboratory test methods when installing and securing a vibration sensor on the handle, hand, and forearm of the operator. Determining the effectiveness of the entire product model should be considered when designing and manufacturing new hand protection against vibration.

CIRA: Class imbalance resilient adaptive Gaussian process classifier

The problem of class imbalance is pervasive across various real-world applications, resulting in machine learning classifiers exhibiting bias towards majority classes. Algorithm-level balancing approaches adapt the machine learning algorithms to learn from imbalanced datasets while preserving the data’s original distribution. The Gaussian process classifier is a powerful machine learning classification algorithm, however, as with other standard classifiers, its classification performance could be exacerbated by class imbalance. In this work, we propose the Class Imbalance Resilient Adaptive Gaussian process classifier (CIRA), an algorithm-level adaptation of the binary Gaussian process classifier to alleviate the class imbalance. To the best of our knowledge, the proposed algorithm (CIRA) is the first adaptive method for the Gaussian process classifier to handle unbalanced data. The proposed CIRA algorithm consists of two balancing modifications to the original classifier. The first modification balances the posterior mean approximation to learn a more balanced decision boundary between the majority and minority classes. The second modification adopts an asymmetric conditional prediction model to give more emphasis to the minority points during the training process. We conduct extensive experiments and statistical significance tests on forty-two real-world unbalanced datasets. Through the experiments, our proposed CIRA algorithm surpasses six popular data sampling methods with an average of 2.29%, 3.25%, 3.67%, and 1.81% in terms of the Geometric mean, F1-measure, Matthew correlation coefficient, and Area under the receiver operating characteristics curve performance metrics respectively.

The Influence of Gas-Dynamic Non-Stationarity of Air Flow on the Heat Transfer Coefficient in Round and Triangular Straight Pipes with Different Turbulence Intensities

Unsteady gas-dynamic phenomena in pipelines of complex configuration are widespread in heat exchange and power equipment. Therefore, studying the heat transfer level of pulsating air flows in round and triangular pipes with different turbulence intensities is a relevant and significant task for the development of science and technology. The studies were conducted on a laboratory stand based on the thermal anemometry method and an automated system for collecting and processing experimental data. Rectilinear round and triangular pipes with identical cross-sectional areas were used in the work. Flow pulsations from 3 to 15.8 Hz were generated by means of a rotating flap. The turbulence intensity (TI) of the pulsating flows varied from 0.03 to 0.15 by installing stationary flat turbulators. The working medium was air with a temperature of 22 ± 1 °C moving at a speed from 5 to 75 m/s. It was established that the presence of gas-dynamic unsteadiness leads to an increase in the TI by 47–72% in a round pipe and by 36–86% in a triangular pipe. The presence of gas-dynamic unsteadiness causes a heat transfer intensification in a round pipe by 26–35.5% and by 24–36% in a triangular pipe. It was shown that a significant increase in the TI of pulsating flows leads to an increase in the heat transfer coefficient by 11–16% in a round pipe and a decrease in the heat transfer coefficient by 7–24% in a triangular pipe. The obtained results can be used in the design of heat exchangers and gas exchange systems in power machines, as well as in the creation of devices and apparatuses of pulse action.

Neural force functional for non-equilibrium many-body colloidal systems

Abstract We combine power functional theory and machine learning to study non-equilibrium overdamped many-body systems of colloidal particles at the level of one-body fields. We first sample in steady state the one-body fields relevant for the dynamics from computer simulations of Brownian particles under the influence of randomly generated external fields. A neural network is then trained with this data to represent locally in space the formally exact functional mapping from the one-body density and velocity profiles to the one-body internal force field. The trained network is used to analyse the non-equilibrium superadiabatic force field and the transport coefficients such as shear and bulk viscosities. Due to the local learning approach, the network can be applied to systems much larger than the original simulation box in which the one-body fields are sampled. Complemented with the exact non-equilibrium one-body force balance equation and a continuity equation, the network yields viable predictions of the dynamics in time-dependent situations. Even though training is based on steady states only, the predicted dynamics is in good agreement with simulation results. A neural dynamical density functional theory can be straightforwardly implemented as a limiting case in which the internal force field is that of an equilibrium system. The framework is general and directly applicable to other many-body systems of interacting particles following Brownian dynamics.

Prediction of transfusion risk after total knee arthroplasty: use of a machine learning algorithm

IntroductionTotal knee arthroplasty (TKA) carries a significant hemorrhagic risk, with a non-negligible rate of postoperative transfusions. The blood-sparing strategy has evolved to reduce blood loss after TKA by identifying the patient's risk factors preoperatively. In practice, a blood count is often performed postoperatively but rarely altering the patient's subsequent management. This study aimed to identify the preoperative variables associated with hemorrhagic risk, enabling the creation of a machine-learning model predictive of transfusion risk after total knee arthroplasty and the need for a complete blood count. HypothesisBased on preoperative data, a powerful machine learning predictive model can be constructed to estimate the risk of transfusion after total knee arthroplasty. Material and methodsThis retrospective single-centre study included 774 total knee arthroplasties (TKA) operated between January 2020 and March 2023. Twenty-five preoperative variables were integrated into the machine learning model and filtered by a recursive feature elimination algorithm. The most predictive variables were selected and used to construct a gradient-boosting machine algorithm to define the overall postoperative transfusion risk model. Two groups were formed of patients transfused and not transfused after TKA. Odds ratios were determined, and the area under the curve evaluated the model's performance. ResultsOf the 774 TKA surgery patients, 100 were transfused postoperatively (12.9%). The machine learning predictive model included five variables: age, body mass index, tranexamic acid administration, preoperative hemoglobin level, and platelet count. The overall performance was good with an area under the curve of 0.97 [95% CI 0.921–1], sensitivity of 94.4% [95% CI 91.2–97.6], and specificity of 85.4% [95% CI 80.6–90.2]. The tool developed to assess the risk of blood transfusion after TKA is available at https://arthrorisk.com. ConclusionThe risk of postoperative transfusion after total knee arthroplasty can be predicted by a model that identifies patients at low, moderate, or high risk based on five preoperative variables. This machine learning tool is available on a web platform that is accessible to all, easy to use, and has a high prediction performance. The model aims to limit the need for routine check-ups, depending on the risk presented by the patient. Level of evidenceII; diagnostic study

Performance Comparison of Support Vector Machines, AdaBoost, and Random Forest for Sentiment Text Analysis and Classification

In sentiment analysis, text analysis becomes an important process to derive useful information from the unstructured data. In this work, we study the performance of three advanced machine learning algorithms, Support Vector Machines (SVM), Random Forest, and AdaBoost, for a specific sentiment classification task. Each classifier was trained and evaluated on principal metrics such as Area Under the Curve (AUC), Classification Accuracy (CA), F1-Score, Precision, and Recall using the Bag of Words model for feature extraction. Those results show that the Random Forest approach beat both SVM and AdaBoost, with an AUC of 0.988, CA = 0.915, and F1-Score = 0.915 SVM demonstrated moderate performance with an AUC of 0.939 and an F1-Score of 0.845, while AdaBoost exhibited the worst performance in all metrics based on that ensemble model-based classifiers for data change predictions. Random Forest may thus be a powerful machine learning technique to implement for sentiment analysis in text.

Ionic species representations for materials informatics

High-dimensional representations of the elements have become common within the field of materials informatics to build useful, structure-agnostic models for the chemistry of materials. However, the characteristics of elements change when they adopt a given oxidation state, with distinct structural preferences and physical properties. We explore several methods for developing embedding vectors of elements decorated with oxidation states. Graphs generated from 110 160 crystals are used to train representations of 84 elements that form 336 species. Clustering these learned representations of ionic species in low-dimensional space reproduces expected chemical heuristics, particularly the separation of cations from anions. We show that these representations have enhanced expressive power for property prediction tasks involving inorganic compounds. We expect that ionic representations, necessary for the description of mixed valence and complex magnetic systems, will support more powerful machine learning models for materials.

FairColor: An efficient algorithm for the Balanced and Fair Reviewer Assignment Problem

As the volume of submitted papers continues to rise, ensuring a fair and accurate assignment of manuscripts to reviewers has become increasingly critical for academic conference organizers. Given the paper-reviewer similarity scores, this study introduces the Balanced and Fair Reviewer Assignment Problem (BFRAP), which aims to maximize the overall similarity score (efficiency) and the minimum paper score (fairness) subject to coverage, load balance, and fairness constraints. Addressing the challenges posed by these constraints, we conduct a theoretical investigation into the threshold conditions for the problem’s feasibility and optimality. To facilitate this investigation, we establish a connection between BFRAP, defined over m reviewers, and the Equitable m-Coloring Problem. Building on this theoretical foundation, we propose FairColor, an algorithm designed to retrieve fair and efficient assignments. We compare FairColor to Fairflow and FairIR, two state-of-the-art algorithms designed to find fair assignments under similar constraints. Empirical experiments were conducted on four real and two synthetic datasets involving (paper, reviewer) matching scores ranging from (100,100) to (10124,5880). Results demonstrate that FairColor is able to find efficient and fair assignments quickly compared to Fairflow and FairIR. Notably, in the largest instance involving 10,124 manuscripts and 5680 reviewers, FairColor retrieves fair and efficient assignments in just 67.64 s. This starkly contrasts both other methods, which require significantly longer computation times (45 min for Fairflow and 3 h 24 min for FairIR), even on more powerful machines. These results underscore FairColor as a promising alternative to current state-of-the-art assignment techniques.