The task of studying and revising can be perceived as challenging due to the intricate nature of concepts presented in dense textbook material. Furthermore, it has been observed that students allocate as much as 50% o...
The task of studying and revising can be perceived as challenging due to the intricate nature of concepts presented in dense textbook material. Furthermore, it has been observed that students allocate as much as 50% of their study time towards study techniques that are deemed less effective [1]. Thus, encouraging students to employ efficient methods is important. Additionally, numerous studies have demonstrated that flashcards are an exceptionally efficient method for enhancing learning outcomes. [2]. To tackle this issue, our research indicates the utilization of various advanced Natural Language Processing models. These models would be combined to develop a multi-stage transformer architecture that enables the automated creation of flashcards from text containing technically rich literature.
Estimating a person's personality is essential in understanding how they manage stress, lead, communicate, collaborate, and influence others. One of the critical factors that affect how people interact with their ...
Estimating a person's personality is essential in understanding how they manage stress, lead, communicate, collaborate, and influence others. One of the critical factors that affect how people interact with their environment is their personality. This project aims to utilize machine learning and also natural language processing techniques to predict the personality of students based on a questionnaire. The system analyzes the user's personality in the database, which already records personality traits. The system then identifies the user's personality based on the MBTI model, a widely used personality model. This project's significance lies in the capability of the system to aid students, teachers, and businesses in identifying students' personalities to plan their actions better. The ML and NLP techniques implemented in this system can categorize or anticipate students' personalities, thus assisting businesses in their hiring process.
Machine learning method is efficient and effective in detecting DDoS attacks, but it all begins from identifying and selecting their important features. This paper presents an implementation of feature selection for D...
Machine learning method is efficient and effective in detecting DDoS attacks, but it all begins from identifying and selecting their important features. This paper presents an implementation of feature selection for DDoS detection based on Random Forest method. In our implementation, we use a LOIC software flood DDoS requests to a target computer, then control the target to extract the features from the captured IP packets, and finally calculate their Gini feature importance and ranking for subsequent feature selection.
The most integral step in computer vision and image processing tasks is the representation of images. Recently, continuous image parameterization using Implicit Neural Representations (INR) has shown a great advantage...
The most integral step in computer vision and image processing tasks is the representation of images. Recently, continuous image parameterization using Implicit Neural Representations (INR) has shown a great advantage over discrete representations due to their spatial invariance. This property immediately finds application in the context of single-image super-resolution (SISR) at an arbitrary scale. However, most super-resolution models, including the INR-based Local Implicit Image Function (LIIF), produce only a single output, failing to address the ill-posedness of SISR. Moreover, these models tend to optimize a mean-squared-error (MSE) based loss function which causes blurring and structural distortion in regions exhibiting a high degree of variance (details). Our work proposes a novel uncertainty-aware self-supervised methodology (U-LIIF) that extends on LIIF, to reduce the blurriness and deals with the ill-posedness of SISR. Our U-LIIF does not require any re-training and yields diversified high-resolution images by leveraging model uncertainty. The efficacy of the proposed method is validated by substantial experiments on various benchmark datasets.
Electrochemical analysis systems are increasingly important for applications like Point-of-Care diagnostics and Lab-on-Package sensing. Herein, we propose a system composed of four passive NFC (Near Field Communicatio...
详细信息
ISBN:
(数字)9798350363517
ISBN:
(纸本)9798350363524
Electrochemical analysis systems are increasingly important for applications like Point-of-Care diagnostics and Lab-on-Package sensing. Herein, we propose a system composed of four passive NFC (Near Field Communication) microchips capable of performing open circuit potentiometry through interactions with commercial smartphones. The multi-chip NFC system can be used to quantify four chemical quantities within a unique liquid sample thanks to a microfluidic circuit which feeds four ad-hoc screen-printed electrodes. In this contribution, preliminary calibration on four analytes is reported. The four target quantities considered in this paper are sodium, potassium, calcium, and pH. The sensing matrix will be human blood and fluids derived from food spoilage Thanks to further implementation with microfluidic channels currently under development, the NFC chemical analyses can facilitate domestic monitoring of patients and rapid, on-site test of pieces of food to check food quality and prevent foodborne illnesses.
The popularity of Machine learning (ML) is increased in the field of weather and climate modelling. Enforcement variety from improved resolution and pre-condition to structure scheme propagation and substitute, and la...
详细信息
ISBN:
(数字)9798350388602
ISBN:
(纸本)9798350388619
The popularity of Machine learning (ML) is increased in the field of weather and climate modelling. Enforcement variety from improved resolution and pre-condition to structure scheme propagation and substitute, and lately even to entire ML climate and climatology predictive models. Linear Structural Equation Models (LSES) have been used, employed and mastered, that for the timed which exceeded 25 years that towards this point have remained dormant. Rather more intensive progress of this domain of knowledge has been observed during the last decade. While proceeding entirely along this line of inquiry, it is most comfortable to use a critical approach in ML applications and the vicinity of weather and climate modelling, at least at the earlier stages.
Federated learning-based strategy can confirm data privacy, as well as other related ones like split learning and differential privacy. When enjoying secured data sharing, we also worry that the data privacy preservin...
Federated learning-based strategy can confirm data privacy, as well as other related ones like split learning and differential privacy. When enjoying secured data sharing, we also worry that the data privacy preserving may harm the model’s overall performance. Given the most recent progress in machine learning and AI methodology, we incorporate self-supervised learning to plug in the federated learning framework and the integrated system can guarantee the model performance and data privacy preservation simultaneously. In the integrated framework, we have different clients to keep their own data, and the data are well separated into the attribute half and the label half, for enhanced privacy, not to mention the additional privacy-preserving skill like differential privacy. Given all the aforementioned components, we can still have the privacy-preserving components equipped model performed superior to or is compatible with the performance without the privacy-preserving components. That is, one client can perform better when adding more information from other clients, without the data sharing in between, and no clients own both the attributes and the corresponding labels. The focused topic is anomaly detection and we pay attention to the imbalanced nature of the data which shows additional challenges to the problem. Given the problem, self-supervised learning is especially useful when obtaining the label information is considered non-trivial. After all, we demonstrate the overall model effectiveness when compared to methods without any federated learning components.
Beamforming based on neural networks (NN) has been proposed as nonlinear beamforming. This method improves the performance with more microphones, but it may not be possible to easily increase the number of microphones...
详细信息
Federated optimization or federated learning (FL) involves optimization of the global model or the server model by minimizing the global loss function which is weighted average of all the local loss functions. The opt...
Federated optimization or federated learning (FL) involves optimization of the global model or the server model by minimizing the global loss function which is weighted average of all the local loss functions. The optimization of the global model requires faster convergence to reduce the number of communication rounds or global iterations which is one of the major challenge in federated optimization. This paper propose FONN which handles this communication overhead in federated optimization by utilizing Nys-Newton, while updating local models. As compared to existing state-of-the-art FL algorithms, SCAFFOLD, GIANT and DONE, utilization of Nys-Newton leads to better convergence and reduction in communication rounds or global iterations while achieving a desired performance from the global model which may be observed from the experimental results on various heterogeneously partitioned datasets.
MRI - Magnetic Resonance Imaging is one of the frequently employed imaging modalities for brain anomaly detection. MRI produces a massive quantity of data, which includes a consecutive set of scans taken at several ti...
详细信息
ISBN:
(数字)9798350396157
ISBN:
(纸本)9798350396164
MRI - Magnetic Resonance Imaging is one of the frequently employed imaging modalities for brain anomaly detection. MRI produces a massive quantity of data, which includes a consecutive set of scans taken at several time instants. Since the existence of brain anomalies that exists examined on every MR sequence, manual brain anomaly detection necessitates anatomical knowledge, costly, laborious, and inaccurate due to human error. Automatic brain anomaly segmentation from 3D Magnetic Resonance Image (MRI) is essential to perform proper diagnosis, monitoring, and treatment planning of the disease. However, owing to the structural difficulties such as hazy boundaries with uneven shapes, precise 3D brain tumor demarcation is interesting. With the recent developments of Deep Learning (DL) models, this research study reviews the recent brain anomaly detection and classification models based on DL approach with 3D MRI data. The existing techniques related to brain tumor segmentation, classification, and validations are reviewed. Every reviewed technique is investigated based on the aim, underlying technique used, dataset, and evaluation parameters. Besides, a comparison table is provided by summarizing the reviewed techniques under several aspects. Finally, a quick analysis of the evaluated techniques' results is conducted to gauge their effectiveness.
暂无评论