ML-based Heterogeneous Container Orchestration Architecture

This paper proposes an implementation of container orchestration to accelerate the migration process from one cloud to another. The use of cloud computing as the infrastructure of a system is in high demand. Cloud computing has been adopted in the use of various applications that are very useful for organizations in establishing a system to reduce effort, cost, and time. The system that runs on the same cloud provider has disadvantages. If the cloud provider goes down, then all of the running services will be affected. Therefore, it is needed the solution to migrate the services to some containers with one controller. In this research, two types of architecture are compared, i.e. container orchestration architecture using Kubernetes and simple container architecture using Docker. We use Kubernetes since it is one of the most popular container orchestrators widely used in big company’s productions. The results show that the container creating time average in the container orchestration architecture using Kubernetes is 26 times faster and the percentage average of resource usage on CPU utilization is 3.6 times smaller compared to simple container architecture using Docker.

This study examines the efficiency of automated testing in container orchestration using Kubernetes as an example. Modern IT environments require rapid, reliable, and scalable application deployment, made possible by advancements in containerization technologies and CI/CD automation. The research is based on an analysis of existing studies. The paper explores the theoretical foundations of container orchestration, including Kubernetes architecture, and the principles of automated testing, encompassing unit, integration, performance, and security testing. Practical aspects of integrating testing into CI/CD processes are presented, with a focus on rolling updates, blue-green, and canary deployments, which help minimize the risk of deploying defective code and reduce downtime. The study also discusses future developments in the field, emphasizing AI/ML integration for failure prediction, improved multi-cluster management, and enhanced security measures. The findings demonstrate that implementing automated testing improves the reliability and efficiency of container orchestration, playing a crucial role in optimizing modern IT infrastructures. The information provided in this study will be of interest to researchers in DevOps, automated testing, and container orchestration, as it contributes to a deeper theoretical understanding and practical optimization of quality assurance processes in distributed systems amid ongoing digital transformation.

The emergence of edge computing has also changed the way contemporary applications are designed, with the focus put on achieving high-resilience, low-latency software programs that are deployed on decentralized and resource-limited systems. Edge-native software is software purposely made to operate at the edge of a network, to allow processing of real-time data, lower bandwidth consumption, and greater fault resiliency. This paradigm is particularly critical to those areas, including autonomous systems, industrial automation, smart cities, and energy infrastructure, that need sub-second responsiveness and system trustworthiness. In this paper, an in-depth design framework of edge-native applications based on microservices architecture is described. Some of the main principles that we discuss are decentralization, data locality, stateless vs. stateful design, redundancy, fault domain isolation and so on. Reference architecture is proposed to support scalable deployments in the event of network partition, and self-healing and management qualities in processing consistent data across heterogeneous edge nodes. We consider powered-up technologies such as lightweight container orchestration (i.e., K3S, Docker), edge communication (i.e., MQTT, gRPC), and Chaos engineering. The performance of the architecture in the latency, fault, and resource stress situations is confirmed by experimental results on the Kubernetes-based edge testbed. The results show that edge-native solutions have the potential to surpass the cloud-based model in measures of latency, throughput, fault recovery, and energy efficiency. We also comment on representative use cases like cars on autopilot and predictive maintenance, and discover such issues as the complexity of orchestration, safety, and managing AI models at the edge

The rise of Microservices architectures has introduced a surge in the demand for containerized software provisioning and orchestration. Kubernetes is a leading container orchestration platform widely adopted for its efficient management of containerized applications within cloud datacenters. However, the proliferation of containerized applications, particularly in resource-constrained environments like Internet of Things (IoT) edge devices, has prompted numerous software vendors to release lightweight Kubernetes-derived platforms, tailored for streamlined deployment in such infrastructures. Despite this, selecting the most suitable distribution for edge-based IoT applications remains a challenge due to a lack of comprehensive studies in this domain. This paper presents a comprehensive model for evaluating the performance of lightweight Container Orchestration Distributions (CODs), incorporating resource usage, control-plane performance, and data-plane performance metrics. Also, we conduct a comparative empirical analysis of four prominent CODs (i.e., k0s, k3s, Microk8s, and Microshift) by focusing on their suitability for edge environments. Our findings reveal that k3s is the overall best-performant COD.

A dose deposition matrix (DDM) prediction method using several voxel features and a machine learning (ML) approach is proposed for plan optimization in radiation therapy. Head and lung cases with the inhomogeneous medium are used as training and testing data. The prediction model is a cascade forward backprop neural network where the input is the features of the voxel, including 1) voxel to body surface distance along the beamlet axis, 2) voxel to beamlet axis distance, 3) voxel density, 4) heterogeneity corrected voxel to body surface distance, 5) heterogeneity corrected voxel to beamlet axis, and (6) the dose of voxel obtained from the pencil beam (PB) algorithm. The output is the predicted voxel dose corresponding to a beamlet. The predicted DDM was used for plan optimization (ML method) and compared with the dose of MC-based plan optimization (MC method) and the dose of pencil beam-based plan optimization (PB method). The mean absolute error (MAE) value was calculated for full volume relative to the dose of the MC method to evaluate the overall dose performance of the final plan. For patient with head tumor, the ML method achieves MAE value 0.49 × 10-4 and PB has MAE 1.86 × 10-4. For patient with lung tumor, the ML method has MAE 1.42 × 10-4 and PB has MAE 3.72 × 10-4. The maximum percentage difference in PTV dose coverage (D98) between ML and MC methods is no more than 1.2% for patient with head tumor, while the difference is larger than 10% using the PB method. For patient with lung tumor, the maximum percentage difference in PTV dose coverage (D98) between ML and MC methods is no more than 2.1%, while the difference is larger than 16% using the PB method. In this work, a reliable DDM prediction method is established for plan optimization by applying several voxel features and the ML approach. The results show that the ML method based on voxel features can obtain plans comparable to the MC method and is better than the PB method in achieving accurate dose to the patient, which is helpful for rapid plan optimization and accurate dose calculation. Establishment of a new machine learning method based on the relationship between the voxel and beamlet features for dose deposition matrix prediction in radiation therapy.

Artificial intelligence has been applied in wildfire science and management since the 1990s, with early applications including neural networks and expert systems. Since then, the field has rapidly progressed congruently with the wide adoption of machine learning (ML) methods in the environmental sciences. Here, we present a scoping review of ML applications in wildfire science and management. Our overall objective is to improve awareness of ML methods among wildfire researchers and managers, as well as illustrate the diverse and challenging range of problems in wildfire science available to ML data scientists. To that end, we first present an overview of popular ML approaches used in wildfire science to date and then review the use of ML in wildfire science as broadly categorized into six problem domains, including (i) fuels characterization, fire detection, and mapping; (ii) fire weather and climate change; (iii) fire occurrence, susceptibility, and risk; (iv) fire behavior prediction; (v) fire effects; and (vi) fire management. Furthermore, we discuss the advantages and limitations of various ML approaches relating to data size, computational requirements, generalizability, and interpretability, as well as identify opportunities for future advances in the science and management of wildfires within a data science context. In total, to the end of 2019, we identified 300 relevant publications in which the most frequently used ML methods across problem domains included random forests, MaxEnt, artificial neural networks, decision trees, support vector machines, and genetic algorithms. As such, there exists opportunities to apply more current ML methods — including deep learning and agent-based learning — in the wildfire sciences, especially in instances involving very large multivariate datasets. We must recognize, however, that despite the ability of ML models to learn on their own, expertise in wildfire science is necessary to ensure realistic modelling of fire processes across multiple scales, while the complexity of some ML methods such as deep learning requires a dedicated and sophisticated knowledge of their application. Finally, we stress that the wildfire research and management communities play an active role in providing relevant, high-quality, and freely available wildfire data for use by practitioners of ML methods.

The all-atomic full-dimensional-level simulations of nonadiabatic molecular dynamics (NAMD) in large realistic systems has received high research interest in recent years. However, such NAMD simulations normally generate an enormous amount of time-dependent high-dimensional data, leading to a significant challenge in result analyses. Based on unsupervised machine learning (ML) methods, considerable efforts were devoted to developing novel and easy-to-use analysis tools for the identification of photoinduced reaction channels and the comprehensive understanding of complicated molecular motions in NAMD simulations. Here, we tried to survey recent advances in this field, particularly to focus on how to use unsupervised ML methods to analyze the trajectory-based NAMD simulation results. Our purpose is to offer a comprehensive discussion on several essential components of this analysis protocol, including the selection of ML methods, the construction of molecular descriptors, the establishment of analytical frameworks, their advantages and limitations, and persistent challenges.

Developing Theory Using Machine Learning Methods

Machine learning guided design of functional materials with targeted properties

How Magic a Bullet Is Machine Learning for Credit Analysis? An Exploration with FinTech Lending Data

Parameter importance assessment improves efficacy of machine learning methods for predicting snow avalanche sites in Leh-Manali Highway, India

Machine learning (ML) methods are attracting considerable attention among academics in the field of finance. However, it is commonly believed that ML has not transformed the asset management industry to the same extent as other sectors. This survey focuses on the ML methods and empirical results available in the literature that matter most for active portfolio management. ML has asset management applications for signal generation, portfolio construction, and trade execution, and promising findings have been reported. Reinforcement learning (RL), in particular, is expected to play a more significant role in the industry. Nevertheless, the performance of a sample of active exchange-traded funds (ETF) that use ML in their investments tends to be mixed. Overall, ML techniques show great promise for active portfolio management, but investors should be cautioned against their main potential pitfalls. 
\n
\n

Prediction of chronic kidney disease (CKD) has emerged as a useful technique for early detection of at-risk persons and the introduction of appropriate management strategies. Machine learning and data-driven methods have been used in predictive modeling to examine massive databases of patient demographics, medical histories, test findings, and genetic information. These cutting-edge methods allow for the profiling of high-risk patients and the tailoring of healthcare administration approaches. Patient outcomes, complication rates, and healthcare system efficiency may all benefit greatly from CKD screening and prediction. Responsible use of CKD prediction algorithms, however, requires resolving issues with data availability, integration, and ethics. The area of medicine has benefited greatly from the use of Machine Learning (ML) methods, which have played an increasingly central role in illness prediction. In this study, we use a strategy that makes use of ML methods to construct effective tools for predicting the development of CKD. Multiple ML models are trained, and their results are compared using a variety of criteria. We applied five ML methods such as logistic regression (LR), Decision tree (DT), random forest (RF), support vector machine (SVM), and k-nearest neighbor (KNN). The LR and KNN have the highest accuracy with 99%.

Background Recently, the need for artificial intelligence (AI) and machine learning (ML) methods in drug development and research is gaining high concern and more grounds. Moreover, providing pharmaceutical and related schools with non-commercial, free-to-use programming languages, software and tools is becoming an unavoidable need. The R programming language can be easily used, through the correct and simplified codes and packages, in conducting unsupervised ML methods, such as principal component analysis (PCA) and hierarchical clustering analysis (HCA), after calculating relevant descriptors of drugs and molecules. Objective The objective of this study was to assess the enhancement of non-computer sciences-based students’ perception of the use of machine learning methods such as PCA and HCA using R-programming in drug formulation. Results Undergraduate students were taught to use R program to derive PCA distinguishable plots such as score, loading and scree, in addition to HCA dendrograms, in the context of developing new pharmaceutical formulations. Surveys conducted pre- and post-teaching the course proved that implementation of such ML methods can help in better understanding and exploring the data, in order to derive meaningful conclusions, and make informed decisions that help develop pharmaceutical formulations of premium quality, with minimal resources consumption. Conclusion We hereby report the easy use of R-programming in applications and activities that introduce undergraduate Pharmaceutical Engineering and Biotechnology students to ML methods. Student surveys showed better student satisfaction and understanding of AI applications in solving pharmaceutical problems. We claim that these students and early_career researchers, who are non-specialists in computer science, can utilize R-programming to perform important pharmaceutical applications through the step-by-step guide and codes provided in this article. Graphical Abstract

Cloud computing has significantly transformed the way organizations utilize computing resources by offering scalability, flexibility, and cost-effectiveness. Energy consumption has become a major research concern with the growing demand for cloud services these days. Adaptive energy optimization techniques are very much essential to minimize energy usage across cloud servers. The main objective of this research is to conserve energy across heterogeneous cloud servers in datacentres. The proposed adaptive energy optimization framework uses container orchestration utilizing Kubernetes framework by sharing the host system kernel and reducing resource overhead to maximize energy efficiency. The work also incorporates smart algorithms, including Ant Colony Optimization (ACO), Swarm Intelligence, and Dynamic Voltage and Frequency Scaling (DVFS) to enable the real-time, energy cognizant distribution of workloads according to the current demand, avoiding unnecessary consumption of resources. Kubernetes container orchestration to optimize energy usage, facilitating effective scaling and thus simplifying the management of applications. Two novel components, the Adaptive Resource Adjustment Algorithm (ARAA) and the Automated Resource-Aware Kubernetes Lifecycle Management (ARKLM), work together to dynamically adjust resource allocation and deactivate idle containers within the Kubernetes environment. Experimental results confirm the framework’s effectiveness, demonstrating that the DVFS algorithm achieved a low 0.94% average CPU usage, while the ACO algorithm delivered the fastest execution time. This dual-focus on energy efficiency and speed makes the framework a scalable and sustainable solution for modern cloud operations, aligning with UNSustainable Development Goals (7, 9, 12, 13, 17) by enabling energy-efficient, low-carbon, and collaborative cloud operations through intelligent automation.