Supplier evaluation has a crucial role in maintaining efficiency in the food industry supply chain. Machine learning approaches can be employed to formulate models aimed at analyzing and evaluating supplier performanc...
详细信息
Supplier evaluation has a crucial role in maintaining efficiency in the food industry supply chain. Machine learning approaches can be employed to formulate models aimed at analyzing and evaluating supplier performance. Previous research has successfully designed decision tree and neural network models for assessing suppliers in the food industry with accuracies of 84.2% and 92.8% separately. Recognizing the opportunity to improve the model's performance, this study aims to advancing the machine learning models accuracy for analyzing and evaluating suppliers in the food industry. Two main models are proposed to enhance model accuracy: ensemble methods and support vector machine. This research has successfully designed a supplier evaluation model and demonstrated that the ensemble method - gradient boosting model outperforms other ensemble methods and support vector machine which is achieved a accuracy of 93.6% on a cross-validation dataset. The development of a dashboard is required to implement the supplier evaluation model using machine learning, facilitating decision-makers in evaluating and controlling supplier performance.
The objective of the newest manufacturing paradigm, known as 'Industry 5.0,'is to combine human intelligence with cutting-edge technology like the Internet of Things (IoT), artificial intelligence (AI), and ro...
详细信息
The concept of Digital Twin has been widely used by researchers to represent physical entities in computer-generated reality in the metaverse. In this research, a novel concept of 'Mobile Twin' is coined. Mobi...
详细信息
This paper applies ant colony optimization (ACO) algorithm for the dual-pin flying probe circuit board inspection optimal path searching problem. First, the proposed approach creates a representation for circuit inspe...
详细信息
The Plasmodium parasite, which causes malaria, is an acute fever illness that infects people when a female Anopheles mosquito bites them. It is predicted that malaria would claim 619,000 lives in 2021, with 96% of tho...
详细信息
ISBN:
(数字)9798331529376
ISBN:
(纸本)9798331529383
The Plasmodium parasite, which causes malaria, is an acute fever illness that infects people when a female Anopheles mosquito bites them. It is predicted that malaria would claim 619,000 lives in 2021, with 96% of those deaths occurring in the African continent. We can achieve this by using a microscope to examine thick and thin blood smears. The proficiency of a microscope examiner is crucial for doing microscopic examinations. Consider how time-consuming, ineffective, and costly it would be to examine thousands of malaria cases. Consequently, Creating an automated method for detecting malaria parasites is the aim of this study. We employ a MobileNetV2 pretrained model with CNN technology. Because it has been trained on dozens or even millions of data points, this pretrained model is incredibly light but dependable. There are two main benefits of automatic malaria parasite detection: firstly, it can offer a more accurate diagnosis, particularly in locations with limited resources; secondly, it lowers diagnostic expenses. The optimizer utilizes Adam Weight, the criteria uses NLLLoss, and the model is trained using 32 for batch_size. In the fourteenth epoch, we obtained the maximum accuracy score of 96.26% based on the training data. The outcomes of the predictions demonstrate how excellent this score is. EfficienceNet, DenseNet, AlexNet, and other pretrained models are among the alternatives that scientists are advised to try training with.
Water resources are inevitable for human survival but untreated wastewater harms the environment. Thus, ongoing monitoring of water quality is necessary to identify pollution sources and prevent further damage. For su...
详细信息
The field of dermatology faces considerable challenges when it comes to early detection of skin cancer. Our study focused on using different datasets, including original data, augmented data, and SMOTE oversampled dat...
详细信息
Tuberculosis is a legal infectious lung disease, a chronic infectious disease caused by tuberculosis. The diagnosis of the confirmed person is usually judged by chest X-ray images, and the sputum smear is used to dete...
详细信息
In general, public or private organizations or companies have used information-based technology as a support to improve business performance to be more effective and efficient in order to achieve a company's busin...
详细信息
Real-world practical systems inherently exhibit non-linearities in their dynamics. Also, it is known that a time-varying delay exists in the system state or input-output. Combined, it affects the stability of the clos...
详细信息
Real-world practical systems inherently exhibit non-linearities in their dynamics. Also, it is known that a time-varying delay exists in the system state or input-output. Combined, it affects the stability of the closed-loop system. It also increases the complexity of the controller design. In-depth controller design research on the class of Nonlinear Systems with Time-Varying Delay (NSTVD) has been the focus of the control community for many years. However, there is a lack of Systematic Literature Review (SLR) and classifications of the papers on this topic. This paper aims to review controller design utilizing a neural network model for the class of NSTVD systems. The study employs Kitchenham’s SLR method to gather, analyze and synthesize published papers from reliable databases between 2017 and 2021. The bibliometric analysis for the selected 38 papers reveals the prolific authors, countries, affiliations, publishers, co-authorship network, co-occurrences of keywords, and ten most-cited papers. Finally, this paper developed a conceptual map outlining six multi-layered findings: the addressed problem, control design method, nonlinear system properties, time-varying delay properties, system constraint properties, and actuator limit properties. A brief qualitative analysis of the ten most-cited papers is performed based on the map. The findings highlighted that the proposed methods have shown encouraging results in the simulation domain and can be used as a source of inspiration for future studies and implementation of the neural controller design of the NSTVD system.
暂无评论