In the past decade, we have seen many patients and healthcare problems. Due to this, patients find difficulty choosing doctors according to their disease. Several machinelearning (ML) based techniques already exist t...
详细信息
ISBN:
(纸本)9781450397964
In the past decade, we have seen many patients and healthcare problems. Due to this, patients find difficulty choosing doctors according to their disease. Several machinelearning (ML) based techniques already exist to predict doctors based on patient's health conditions. However, it is essential to accurately recommend doctors to patients with low errors based on patients' health conditions. Therefore, we propose a method that assigns quantitative importance (weight) to each feature using an ML technique. Moreover, we offer a framework to recommend doctors based on the similarity score and doctor's skill score, which utilizes weight prediction to enhance operational efficiency. Additionally, on real-world datasets, the effectiveness of the proposed framework is demonstrated empirically by lowering the average loss by roughly 34% and 3% as compared to Convolutional Neural Network (CNN) and Support Vector machine (SVM), respectively. The outcome demonstrates that the algorithm can efficiently recommend doctors to patients compared to state-of-the-art techniques. This analysis technique aid patients in opting for the right doctor.
In this work, a novel technique for detecting extra-terrestrial bodies using the transit method, with the aim of improving traditional algorithmic strategies in astronomy through machinelearning algorithms. The poten...
详细信息
Traditional methods for kick prediction rely on experience and basic physical models, but they struggle to maintain high accuracy under complex well conditions and varying geological environments. Moreover, convention...
详细信息
ISBN:
(纸本)9789819770069;9789819770076
Traditional methods for kick prediction rely on experience and basic physical models, but they struggle to maintain high accuracy under complex well conditions and varying geological environments. Moreover, conventional approaches often fail to fully leverage big data resources. Therefore, this study aims to improve the accuracy of kick prediction through machinelearning models, especially by adopting an innovative feature fusion strategy to optimize the prediction model. We evaluated the comprehensive performance of six different machinelearning models: GBM, LightGBM, CatBoost, XGBoost, Random Forest, and AdaBoost. Based on these evaluations, we proposed a feature score weighted fusion method. This method weights the importance scores of features based on the accuracy of each model, selecting the features that have the most significant impact on the prediction results. Experimental results show that the model utilizing the feature fusion strategy performs better on the test set than any single base model. Specifically, the GBM model, optimized through feature fusion, achieved better values in accuracy, F1 score, and F2 score, proving the effectiveness of this feature fusion strategy in enhancing the accuracy of kick predictions during well control operations. Furthermore, this study compared the performance of different models, providing valuable insights for kick prediction in well control operations. This research not only demonstrates the potential of feature fusion strategies in improving the accuracy of kick prediction but also offers new insights for the future application of machinelearning technologies in the field of oilfield data analysis.
Cloud computing stands as one of the most pervasive technological innovations, providing computing services, including IT infrastructure, over the internet. This technology alone saves many small and medium-sized IT c...
详细信息
With the development of smart grid, the path planning of digital transmission lines has become an important research direction. The path planning of transmission line is to determine the direction of transmission line...
详细信息
Sensing and onboard-processing capabilities of next-generation spacecraft continue to evolve. Enabled by advances in avionic systems, large amounts of data can be collected and stored on orbit. Nevertheless, loss of s...
详细信息
ISBN:
(纸本)9798350384543;9798350384536
Sensing and onboard-processing capabilities of next-generation spacecraft continue to evolve. Enabled by advances in avionic systems, large amounts of data can be collected and stored on orbit. Nevertheless, loss of signal, communication delays, and limited downlink rates remain a bottleneck for delivering data to ground stations or between satellites. This research investigates a multistage image-processing pipeline and demonstrates rapid collection, detection, and transmission of data using the Space Test Program - Houston 7 - Configurable and Autonomous Sensor Processing Research (STP-H7-CASPR) experiment aboard the international Space Station as a case study. machine-learning (ML) models are leveraged to perform intelligent processing and compression of data prior to downlink to maximize available bandwidth. Moreover, to ensure accuracy and preserve data integrity, a fault-tolerant ML framework is employed to increase pipeline reliability. This pipeline fuses the fault-tolerant Resilient TensorFlow framework with ML-based tile classification and the CNNJPEG compression algorithm. This research shows that the imaging pipeline is able to alleviate the impact of limited communication bandwidth by using reliable, autonomous data processing and compression techniques to achieve reduced transfer sizes of essential data. The results highlight the benefits provided by resilient classification and compression including minimized storage use and reduced downlink time. Furthermore, the findings of this research are used to assess the feasibility of such a system for future space missions. The combination of these approaches enables the system to achieve up to a 98.67% reduction in data size and downlink time as well as the capacity to capture imagery up to 75.19x longer time period for a given storage size, respectively, while maintaining reconstruction quality and data integrity.
In this paper, we introduce GraphGPT, a novel machine-learning approach designed for intuitive, logical, and visual education. GraphGPT is inspired by the observation that scientific graphs help learners visualize con...
详细信息
Emotions play a pivotal role in human communication, influencing our interactions and relationships profoundly. Understanding and accurately detecting emotions in multimodal communication is essential for various appl...
详细信息
ISBN:
(纸本)9798350385939;9798350385922
Emotions play a pivotal role in human communication, influencing our interactions and relationships profoundly. Understanding and accurately detecting emotions in multimodal communication is essential for various applications, from human-computer interaction to mental health assessment. This research paper explores a novel approach to emotion detection by leveraging both audio and video modalities in the context of gesture recognition. The requirement for a multimodal approach in our work stems from the limitations of unimodal emotion detection systems. While individual modalities like audio or video have shown promise in emotion recognition, they often fail to capture the full spectrum of emotional cues present in human communication. By integrating these modalities, we aim to achieve more robust and comprehensive emotion recognition results. We present a comprehensive analysis of our proposed system's architecture, experimental methodology, and results, demonstrating its potential to significantly enhance the accuracy and reliability of emotion detection in real-world scenarios. The proposed framework achieves an overall accuracy of 87%, showcasing its effectiveness in addressing the challenge of improving emotion recognition accuracy in multimodal communication.
machinelearning has a crucial role in people's lives. machinelearning can be divided into four parts: supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning. It can help ...
详细信息
The data from remote sensing satellites provide opportunities to acquire information about land at varying resolutions and has been widely used for change detection studies. However, due to the noise interference and ...
详细信息
ISBN:
(纸本)9798400709234
The data from remote sensing satellites provide opportunities to acquire information about land at varying resolutions and has been widely used for change detection studies. However, due to the noise interference and redundancy in remote sensing multispectral images, it is valuable to transform the available spectral channels into a suitable feature space to alleviate noise and reduce redundancy. In this paper, we propose a supervised change detection method for optical aerial images based on deep Siam convolution network, which is mainly composed of two branch network, heuristic feature aggregation module and feature attention module. This method achieves better performance in experiments on remote sensing aerial public datasets. We show that our network can extract feature from two images at a same time, and depend on feature fusion strategy together with attention mechanism to further enrich the difference information. Subsequently, a decoder structure of semantic segmentation is developed to harvest binary change detection result. Experimental results show that our algorithm achieves a comprehensive F1 score of 83.9% on different data sets, which is superior to the traditional algorithm and other deep learning networks, and has a certain degree of effectiveness and robustness.
暂无评论