With the exponential growth of remote sensing data, particularly multispectral imagery, the integration of deep learning techniques such as Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs) ha...
详细信息
ISBN:
(数字)9798350354171
ISBN:
(纸本)9798350354188
With the exponential growth of remote sensing data, particularly multispectral imagery, the integration of deep learning techniques such as Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs) has become relevant for efficient classification tasks. Satellite imagery is procured in the manner of multispectral imagery, which is a collection of a few image layers enlist of the same scene, each of them acquired at a particular wavelength band. Challenges in multispectral images arise owing to the large volume of image data and the data variety due to high dimensionality and number of channels. Land Cover and Terrain classification is the process of identifying, mapping and delineating the surface of the earth into different units. It remains vital for its applications in a variety of fields. It enables environmental monitoring allowing for adaptation to ecosystem variations. Fields like Urban Planning, Disaster Management and Infrastructure development find essential advantages in their planning phases owing to such classification. Deep CNNs, or Convolutional Neural Networks, are advanced machinelearning models renowned for their ability to extract intricate patterns from visual data, crucial in tasks like image recognition and classification. The reason behind the application of deep CNNs arises from their unparalleled computing efficiency. Through our work we try to explore the utilization of above mentioned deep learning and machinelearning techniques in the classification of land cover, focusing on both multispectral imagery and provide an overview of the existing literature, examining the methodologies employed and the datasets utilized and the processing and handling of these large datasets. Additionally, we introduce various data sources and datasets commonly utilized in these studies, shedding light on the advancements, challenges, and future directions in the field of remote sensing-based land use and land cover classification.
Text mining of power system equipment has important applications in fault diagnosis. With the improvement of intelligence and informatization level of power system equipment, a large amount of device text data is reco...
Text mining of power system equipment has important applications in fault diagnosis. With the improvement of intelligence and informatization level of power system equipment, a large amount of device text data is recorded and stored, which contains rich fault information. Through text mining technology, valuable information can be extracted from these data for equipment fault diagnosis and prediction. This paper exploited various model parameter structures such as LSTM, universal BERT, power text BERT, and Improved BERT on text exaction performance in the context of power systems, and compared the performance differences between these models under small sample tasks. The solutions are evaluated through experiments and the numerical results are provided and discussed.
With the development of IT technology and the popularization of Internet applications, the scale and value of information carried and disseminated on the Internet are increasing. It has become one of the most importan...
详细信息
This study explores state-of-the-art advanced ensemble learning methodologies for predictive modeling in marathon running times. The research emphases on enhancing the precision and reliability of marathon time predic...
详细信息
BP neural network, or back propagation neural network, is a machinelearning algorithm widely used in patternrecognition, function approximation, signal processing and other fields. In recent years, with the expansio...
BP neural network, or back propagation neural network, is a machinelearning algorithm widely used in patternrecognition, function approximation, signal processing and other fields. In recent years, with the expansion of power grid scale and the increase of complexity, BP neural network plays an increasingly important role in the three-dimensional processing of power grid. This paper mainly discusses the application of BP neural network in intelligent assistance, evaluation and optimization of power grid.
In the healthcare industry, machinelearning (ML) plays a crucial role in disease prediction. A patient must go through a series of tests before a condition can be diagnosed. However, using machinelearning techniques...
详细信息
In the healthcare industry, machinelearning (ML) plays a crucial role in disease prediction. A patient must go through a series of tests before a condition can be diagnosed. However, using machinelearning techniques, the number of tests can be reduced. This simplified test has a significant impact on both time and performance. Early patient care has benefited from sound medical data analysis due to the growing amount of data generated by the medical and healthcare sectors. With the help of disease data, massive amounts of medical data can be mined for hidden pattern information. With a focus on heart diseases, this study evaluates and suggests a heart disease prediction based on the patient's symptoms using machinelearning techniques such as SVM, MLR, and RF algorithms. The proposed method outperforms those currently in use in terms of accuracy, forecast speed, and consistency of outcomes. It is also appropriate to classify lung cancers using trained datasets for accurate identification.
machinelearning and AI-based automated systems are gaining increasing attention for real-time intelligent applications by virtue of a superior co-ordination between the software and the hardware within these systems....
详细信息
ISBN:
(纸本)9781665419130
machinelearning and AI-based automated systems are gaining increasing attention for real-time intelligent applications by virtue of a superior co-ordination between the software and the hardware within these systems. Although the majority of the automated systems are implementing Convolutional neural networks (CNNs), and Deep Neural Networks (DNNs) on the hardware with impressive accuracy, a significant amount of cost is associated with data movement in these platforms. Recent advancements in processing-in-memory (PIM), a non-von Neumann computing paradigm, have proven to be very effective in minimizing data communication overheads by performing computations within the memory chip. However, these devices are primarily designed as inference engines and therefore have not been adequately investigated for real-time learning capabilities for applications in changing environments. In this work, we introduce uPIM, a PIM architecture that supports a Generative Adversarial Network (GAN)-based performance-aware online learning model for updating the weights with minimal overheads. Our hardware-software co-design approach exhibits superior performance and efficiency in real-time applications like Autonomous Navigation Systems (ANS) by leveraging massive data-level parallelism and ultra-low data movement latency. The evaluations are performed on multiple state-of-the-art deep learning networks like LeNet, AlexNet, ResNet18, 34, 50 on the German Traffic Sign recognition Benchmark (GTSRB) dataset and the Belgium Traffic Sign dataset (BTSD) with several data-precisions. The proposed performance-aware, quantization-friendly online learning based PIM architecture achieves an average accuracy of 72% for GTSRB and 83.4% for BTSD dataset under varying environment for CNNs implemented for Traffic Sign recognition (TSR) with 8-bit fixed point data-precision.
The rise of measuring, computing, and storage capabilities in modern information systems has led to vast amounts of data in various fields of human activities. Various algorithms for machinelearning have been develop...
详细信息
ISBN:
(数字)9798350372120
ISBN:
(纸本)9798350372137
The rise of measuring, computing, and storage capabilities in modern information systems has led to vast amounts of data in various fields of human activities. Various algorithms for machinelearning have been developed to get valuable insights from this raw data. Tuning the hyperparameters of each algorithm is necessary. Finding the right algorithm and manually tuning its hyperparameters for a given task can be very time-consuming. The methods for automating this search are within the scope of AutoML (automated machinelearning). The wide use of wearable devices such as smartwatches and fitness trackers enables the gathering of various information about the physical activities of the wearer. This information, in turn, can be used to produce useful insights, such as recognizing activities being performed. AutoML can automate the process of building models for this task, but the performance of the resulting models can be very different depending on the AutoML library that performed the task. In this work, we develop an AutoML framework that unites several commonly used AutoML libraries and enables the use of common constraints and metrics. The developed framework is then used to build models for the task of human activity recognition by processing a dataset containing information collected from smart devices worn by users.
This study aims to analyze pattern second year student learning outcome in the Information Systems study program at Telkom University. Telkom University has a Center for e-learning and Open Education (CeLOE) LMS which...
This study aims to analyze pattern second year student learning outcome in the Information Systems study program at Telkom University. Telkom University has a Center for e-learning and Open Education (CeLOE) LMS which is a platform for student learning when online classes. However, the achievement of student learning outcomes has not yet been achieved maximum. This research uses process mining algorithms, specifically the process cube, to analyze event log data from CeLOE LMS. The formulation of the problem of this research itself includes the pattern of student learning path, the best pattern in achieving learning outcomes in the Accounting and Management Systems course Finance as well as Business Process Engineering course, and the performance of the teaching process for lecturers. The purpose of this research is to find out activity pattern student learning, find the best patterns in achieving learning outcomes, and evaluate teaching process performance. This research has limitations on the data obtained from odd semester 2022/2023 and even 2021/2022 and focus on second year students of Information Systems. The research results show that students need to do activities such as doing assignments, attending virtual meetings, doing quizzes, and accessing material learning to achieve pass status, while students who fail just do view action only on the activity. It is important for students to do assignments and quizzes well. In the course Accounting Systems and Management Finance as well as Business Process Engineering, the best pattern of learning includes doing assignments and quizzes simultaneously in appropriate time. This research provides benefit for science, research objects (second year students of Information Systems), and writers itself. Through this research, it is expected to provide benefit for scientific development in the field of process mining and education, as well as give contribution to repair quality learning and student learning outcomes.
machinelearning is used in many supervised applications, nevertheless a learning process needs a sufficiently large amount of annotated data to build a model able to generalize. The lack of such annotated data is a b...
详细信息
暂无评论