Multi-class and multi-label classification on noisy call transcript data generated by speech-to-text (STT) systems is challenging due to the different human accents and transcription errors. The multi-labeling task is...
详细信息
ISBN:
(纸本)9783031477140;9783031477157
Multi-class and multi-label classification on noisy call transcript data generated by speech-to-text (STT) systems is challenging due to the different human accents and transcription errors. The multi-labeling task is even more complicated if the data points have only single or no labels. This study has three main contributions to solving these problems: (1) To overcome the labeling problem, we train a multi-class classification model and use a minimal set of manually annotated data to determine a threshold. We obtain a multi-label classifier by utilizing a multi-class classifier with this threshold. (2) To overcome the noise issue, we propose concatenating well-known feature extraction techniques such as word2vec, tf-idf, transformers, and fuzzy embeddings. This combined feature extraction method is more resilient to noise with proper configurations than stand-alone techniques. (3) This is an industry task;we must protect our client's data. Hence to carry out our success on French private client data to benchmark data, we propose a noising pipeline that artificially mimics the observed STT transcription errors. We combined these solutions in an NLP framework, enabling us to achieve state-of-the-art results with fewer resources, such as manually annotated data or multiple GPU utilization.
Bonds are tradable investment instruments that offer yields, representing the promised return on investment. Unlike fixed-interest bonds, bond yields typically fluctuate, so accurate yield predictions are crucial for ...
详细信息
ISBN:
(纸本)9783031671913;9783031671920
Bonds are tradable investment instruments that offer yields, representing the promised return on investment. Unlike fixed-interest bonds, bond yields typically fluctuate, so accurate yield predictions are crucial for investors. These fluctuations may include increase, decrease, and steady yield values, aligning well with the principles of the neutrosophic soft set. In this study, we apply the neutrosophic soft set theory to predict Indonesian bond yields in a multi-attribute time series framework. We consider closing yield, opening yield, and daily amplitude as predictor variables. We achieve shallow low prediction errors through experiments with varied training data ranges and n-order variations. We discover that the lowest values for Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Root Mean Square Error (RMSE) are 0.0436, 0.6462%, and 0.0514, respectively. These errors are achieve when n = 13, with a two-year train data length. These results underscore the efficacy of the neutrosophic soft set in accurately predicting the closing yield of Indonesian bonds.
Advances in next-generation sequencing and in "-omics" technologies enable the characterization of the human gut microbiome. Colorectal cancer (CRC), the third most common cancer worldwide, is caused by gene...
详细信息
ISBN:
(纸本)9783031671944;9783031671951
Advances in next-generation sequencing and in "-omics" technologies enable the characterization of the human gut microbiome. Colorectal cancer (CRC), the third most common cancer worldwide, is caused by genetic mutations, environmental influences, and abnormalities in the gut microbiota. The aim of this study is to identify pathways that influence host metabolism in CRC patients. The CRC-related metagenomic dataset used in this study contains the relative abundance values of 551 pathways calculated for 1262 samples. Here, two different approaches based on the feature grouping reduce the number of features by considering relevant features as groups, eliminate irrelevant features, and perform classification. The recursive cluster elimination with intra-cluster feature elimination (RCE-IFE) approach achieves anAUCof 0.72 using an average of 66.2 features on CRC-associated metagenomics dataset. In these experiments, P163-PWY: L-lysine fermentation to acetate and butanoate and PWY-6151: S-adenosyl-L-methionine cycle I pathways are identified as potential biomarkers associated with CRC. These experiments also reduce the number of features reported by both approaches in P163-PWY: L-lysine fermentation to acetate and butanoate and PWY-6151: Sadenosyl-L-methionine cycle I pathways reported by both approaches are considered possible CRC-related biomarkers. This study contributes to the molecular diagnosis and treatment of colorectal cancer by revealing the pathways associated with CRC. Our results are promising for the study of the gut microbiota and its role in CRC.
Originally, in rough set theory (RST) proposed by Z. Pawlak, approximation of sets is defined on the basis of an indiscernibility relation between objects in some universe of discourse. The problems appear if attribut...
详细信息
ISBN:
(数字)9783031456510
ISBN:
(纸本)9783031456503;9783031456510
Originally, in rough set theory (RST) proposed by Z. Pawlak, approximation of sets is defined on the basis of an indiscernibility relation between objects in some universe of discourse. The problems appear if attribute values describing objects are symbolical (e.g., linguistic terms). Such a situation is natural in human cognition and description of the real world (e.g. in case of medical applications, where diseases are described in natural language terms). We can perfect rough set theory in this area by incorporating ontologies enabling us to add some new, valuable knowledge, which can be used in data analysis, rule generation, reasoning, etc. In the paper, we propose to use ontological graphs in determining approximations of sets as well as we show how ontological graphs change the look at them in case of linguistic medical data.
This study presents a novel approach to creating lightweight corrugated sandwich panels using additive manufacturing, inspired by eco-friendly green sandwiches composed of Oak-tree cupules as core elements and balsa f...
详细信息
ISBN:
(数字)9783031704628
ISBN:
(纸本)9783031704611;9783031704628
This study presents a novel approach to creating lightweight corrugated sandwich panels using additive manufacturing, inspired by eco-friendly green sandwiches composed of Oak-tree cupules as core elements and balsa face sheets. The bio-inspired structures were fabricated using the Multi Jet Fusion method, employing Nylon PA12 GB infused with 40% glass beads. Through experimental evaluation, the quasi-static compressive characteristics and energy absorption behavior of the bio-inspired sandwich panels were examined to analyze the impact of geometrical dimensions on their ability to withstand compressive loads and absorb energy. Results indicate that the printed panels outperformed the bio-based samples of identical dimensions, demonstrating approximately five times greater peak load and energy absorption. Additionally, the performance of the printed panels regarding load capacity and energy absorption was found to be influenced by their geometrical dimensions.
The COVID-19 pandemic has caused significant health, social, and economic disruptions globally, exposing healthcare systems' vulnerabilities and disparities in healthcare access and outcomes. The global response t...
详细信息
ISBN:
(纸本)9783031614149;9783031614156
The COVID-19 pandemic has caused significant health, social, and economic disruptions globally, exposing healthcare systems' vulnerabilities and disparities in healthcare access and outcomes. The global response to the pandemic has included a variety of measures, including public health interventions, social distancing measures, travel restrictions, and vaccine campaigns. Mathematical and computer modeling has played a crucial role in understanding and combatting the pandemic. The Russian war in Ukraine has caused immense difficulties for medical personnel and severely impacted the accessibility and availability of medical care, disrupting the country's COVID-19 vaccination and prevention efforts. The paper aims to assess the impact of the Russian war in Ukraine on the COVID-19 epidemic process in Canada. We used forecasting methods based on statistical machine learning to build a COVID-19 distribution model. Results showed high accuracy in predicting cumulative new cases and deaths in Canada for 30 days. The model was then applied to the first 30 days of the full-scale Russian invasion to Ukraine, and the study concluded that forced migration of Ukrainians to Canada did not have a significant impact on the epidemic's dynamics. The study's experimental results suggest that the developed model can be used in public health practice.
This study introduces an advanced framework for plant disease detection, specifically classifying tomato images into "Early Blight" and "Healthy" categories. Utilizing a fusion of artificial intell...
详细信息
ISBN:
(数字)9789819720538
ISBN:
(纸本)9789819720521;9789819720538
This study introduces an advanced framework for plant disease detection, specifically classifying tomato images into "Early Blight" and "Healthy" categories. Utilizing a fusion of artificial intelligence and computer vision, the research employs the MobileNet architecture enriched with custom convolutional layers for enhanced feature extraction. The model's adaptability to different dataset sizes highlights its robustness, with performance benchmarks indicating up to 100% accuracy using classifiers like Random Forest, SVM, and Gradient Boosting. The framework further leverages ensemble classifiers to refine prediction accuracy, addressing the real-world complexities of variable lighting and environmental conditions. In its entirety, the research offers a scalable, accurate, and systematic approach to automated plant disease detection, with implications for bolstering global food security and sustainable agriculture.
In the realm of tissue engineering, 3D bioprinting has emerged as a cutting-edge methodology, unlocking unprecedented possibilities for the fabrication of tissue-like structures with potential applications in regenera...
详细信息
ISBN:
(纸本)9783031608391;9783031608407
In the realm of tissue engineering, 3D bioprinting has emerged as a cutting-edge methodology, unlocking unprecedented possibilities for the fabrication of tissue-like structures with potential applications in regenerative medicine. The advancement of 3D bioprinting techniques empowers the creation of intricate, organ-like structures that can either replace damaged portions of organs or serve as substitutes for entire organs. Notable successes in in vivo experiments involving 3D bioprinted skin, bone, and bladder underscore the transformative potential of this state-of-the-art technology. The evolving landscape of healthcare underscores the imperative for a personalized therapeutic approach, necessitating innovative strategies in tissue engineering. The precision offered by modern techniques and devices, notably 3D bioprinters, enhances the success of this multidisciplinary scientific endeavor. A primary focus of our research is the development of a method for the production of artificial blood vessels, responding to the high demands in the field of vascular treatment and healing. The crux of artificial blood vessel bioengineering lies in the utilization of meticulously crafted scaffolds, strategically seeded with stem cells that differentiate into somatic cells within human tissue. In pursuit of the optimal scaffold design for blood vessel production, we propose the application of polyethylene glycol (PEG) and polycaprolactone (PCL) polymers. Our results reveal a chemistry that proves to be optimal for this critical task, paving the way for advancements in the bioengineering of artificial blood vessels. This study contributes to the evolving landscape of tissue engineering and underscores the potential of PEG and PCL polymers in pioneering innovative solutions for vascular therapy.
This article examines the relationship between supply chain collaboration, sustainable supply chain and supply chain performance by providing answers to the following research question: how does collaboration among su...
详细信息
ISBN:
(纸本)9783031686276;9783031686283
This article examines the relationship between supply chain collaboration, sustainable supply chain and supply chain performance by providing answers to the following research question: how does collaboration among supply chain companies contribute to the performance of sustainability in the Moroccan industry? The data of this multiple-case study was collected from semi-structured interviews in three case-companies in Morocco. By examining multiple cases, the research aims to capture variations and commonalities, allowing for a more robust analysis of the research question. The findings emphasize how supply chain collaboration and sustainability synergistically reinforce each other, leading to enhanced sustainability outcomes, improved risk management, and greater overall organizational resilience. This research contributes to the understanding of how Moroccan firms leverage collaborative networks to drive sustainability, positioning them at the forefront of sustainable business practices in the global arena. The research relied on qualitative data collected through semi-structured interviews. While this approach provides rich insights, it might not capture quantitative trends or statistical relationships that can be analyzed using more extensive data sets. Managers should recognize the significance of collaboration among supply chain companies. Implementing structured collaboration strategies, such as joint planning, information sharing, and coordinated decision-making, can contribute to the overall performance of sustainable practices.
Existing data trading markets trade source data, which inevitably raises the issue of data ownership, and more recently model trading markets, which trade on machine learning models that guarantee the ownership of the...
详细信息
ISBN:
(纸本)9783031663284;9783031663291
Existing data trading markets trade source data, which inevitably raises the issue of data ownership, and more recently model trading markets, which trade on machine learning models that guarantee the ownership of the data owner, have become increasingly popular. However, there are still difficulties in the assessment of the utility of the model and the distribution of market benefits. In this paper, we design a framework for a model trading market (MDB) that considers the interaction between data providers and broker. We evaluate the utility of the model in a variety of ways, including data quality, privacy protection needs, and the number of data providers, and reveal the dynamic decision-making process of data providers as they trade off the utility of privacy against the utility of data compensation. At the same time, we analyze the optimal allocation compensation strategy of brokers to maximize the welfare of the whole market. The feasibility and effectiveness of the MDB model trading market are demonstrated by numerical results.
暂无评论