检索结果-内蒙古大学图书馆

Predicting Power Output of Solar Photovoltaic Panels Using Machine Learning Techniques

学校读者我要写书评

暂无评论

Predicting Power Output of Solar Photovoltaic Panels Using M...

Artificial Intelligence, Applications, Innovation and Ethics (AI2E), International Conference for

作者： Karimulla Syed Mohammad Ameer Hamza Yousuf Murali Krishna Boddu Sainadh Singh K Mahaboob Shareef Syed Afaq Ahmad Engineering Department University of Technology and Applied science-Shinas Shinas Sultanate of Oman Department of EEE BVRIT Narsapur India Vignan's Lara ITS India Department of Mathematics and Computer Science Modern College of Bussiness & Science Muscat Sultanate of Oman

ISBN: (数字)9798331513733

ISBN: (纸本)9798331513740

The prediction of solar power generation is essential for effective integration of renewable energy into power grids, aiding in grid stability, energy planning, and efficient resource allocation. Due to the inherent variability of solar energy caused by factors like weather patterns, time of day, and seasonal changes, machine learning (ML) has appeared as a powerful tool to improve forecasting accuracy. Solar panels with various tilt angle combinations are set up to collect experimental data. This paper uses supervised regression learner in machine learning application for solar power prediction. In this paper, linear regression and step wise linear regression models are giving fruitful results out of 27 models that are analyzed. A detailed study of model selection is provided, alongside an examination of evaluation metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Means Square Error (RMSE), and regression coefficient R 2 scores. This study shows the effectiveness of Machine Learning (ML) in enhancing short, and medium-term solar power forecasting, supporting more efficient energy management and promoting the scalability of renewable energy systems. For the linear regression algorithm, we obtained regression coefficient (R 2 ) of 1, mean absolute percentage error (MAPE) of 0.7% and for the stepwise linear regression algorithm, we obtained a regression coefficient (R 2 ) of 1, mean absolute percentage error (MAPE) 0.45%.

关键词： Training Renewable energy sources Analytical models Machine learning algorithms Linear regression Machine learning Prediction algorithms Solar panels Forecasting Meteorology

Identifying Critical Tokens for Accurate Predictions in Transformer-based Medical Imaging Models

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Kang, Solha Vankerschaver, Joris Ozbulak, Utku Center for Biosystems and Biotech Data Science Ghent University Global Campus Korea Republic of Department of Applied Mathematics Computer Science and Statistics Ghent University Belgium Department of Electronics and Information Systems Ghent University Belgium

With the advancements in self-supervised learning (SSL), transformer-based computer vision models have recently demonstrated superior results compared to convolutional neural networks (CNNs) and are poised to dominate the field of artificial intelligence (AI)-based medical imaging in the upcoming years. Nevertheless, similar to CNNs, unveiling the decision-making process of transformer-based models remains a challenge. In this work, we take a step towards demystifying the decision-making process of transformer-based medical imaging models and propose "Token Insight", a novel method that identifies the critical tokens that contribute to the prediction made by the model. Our method relies on the principled approach of token discarding native to transformer-based models, requires no additional module, and can be applied to any transformer model. Using the proposed approach, we quantify the importance of each token based on its contribution to the prediction and enable a more nuanced understanding of the model’s decisions. Our experimental results which are showcased on the problem of colonic polyp identification using both supervised and self-supervised pretrained vision transformers indicate that Token Insight contributes to a more transparent and interpretable transformer-based medical imaging model, fostering trust and facilitating broader adoption in clinical settings. Copyright © 2025, The Authors. All rights reserved.

关键词： Self-supervised learning

Blind Equalization Using a Variational Autoencoder With Second Order Volterra Channel Model

学校读者我要写书评

暂无评论

IEEE Transactions on Cognitive Communications and Networking 2025年

作者： Nielsen, Søren Føns Zibar, Darko Schmidt, Mikkel N. Technical University of Denmark Department of Applied Mathematics and Computer Science United Kingdom Technical University of Denmark Department of Electrical and Photonics Engineering United Kingdom

Existing communication hardware is being exerted to its limits to accommodate for the ever increasing internet usage globally. This leads to non-linear distortion in the communication link that requires non-linear equalization techniques to operate the link at a reasonable bit error rate. This paper addresses the challenge of blind non-linear equalization using a variational autoencoder (VAE) with a second-order Volterra channel model. The VAE framework’s costfunction, the evidence lower bound (ELBO), is derived for real-valued constellations and can be evaluated analytically without resorting to sampling techniques. We demonstrate the effectiveness of our approach through simulations on a synthetic Wiener-Hammerstein channel and a simulated intensity modulated direct detection (IM/DD) optical link. The results show significant improvements in equalization performance, compared to a VAE with linear channel assumptions, highlighting the importance of appropriate channel modeling in unsupervised VAE equalizer frameworks. © 2015 IEEE.

关键词： Blind equalization

Kink movement on a periodic background

学校读者我要写书评

暂无评论

Physical Review E 2025年第2期111卷 024203-024203页

作者： Tomasz Dobrowolski Jacek Gatlik Panayotis G. Kevrekidis Department of Computer Physics and Quantum Computing University of the National Education Commission in Krakow Podchorążych 2 30-084 Cracow Poland AGH University of Krakow Faculty of Physics and Applied Computer Science 30-059 Krakow Poland Department of Mathematics and Statistics University of Massachusetts Amherst Massachusetts 01003-4515 USA

The behavior of the kink in the sine-Gordon (sG) model in the presence of periodic inhomogeneity is studied. An ansatz is proposed that allows for the construction of a reliable effective model with two degrees of freedom. Effective models featuring very good agreement with the original field-theoretic partial differential equation are constructed, including in the nonperturbative region and for relativistic velocities. The numerical solutions of the sG model describing the evolution of the kink in the presence of a barrier as well as in the case of a periodic heterogeneity under the potential additional influence of a switched bias current and/or dissipation were obtained. The results of the field equation and the effective models were favorably compared. The effect of the choice of initial conditions in the field model on the agreement of the results with the effective model is also discussed.

关键词： sine-Gordon equation

Self-supervised Benchmark Lottery on ImageNet: Do Marginal Improvements Translate to Improvements on Similar Datasets?

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Ozbulak, Utku Anzaku, Esla Timothy Kang, Solha De Neve, Wesley Vankerschaver, Joris Center for Biosystems and Biotech Data Science Ghent University Global Campus Korea Republic of Department of Electronics and Information Systems Ghent University Belgium Department of Applied Mathematics Computer Science and Statistics Ghent University Belgium

Machine learning (ML) research strongly relies on benchmarks in order to determine the relative effectiveness of newly proposed models. Recently, a number of prominent research effort argued that a number of models that improve the state-of-the-art by a small margin tend to do so by winning what they call a "benchmark lottery". An important benchmark in the field of machine learning and computer vision is the ImageNet where newly proposed models are often showcased based on their performance on this dataset. Given the large number of self-supervised learning (SSL) frameworks that has been proposed in the past couple of years each coming with marginal improvements on the ImageNet dataset, in this work, we evaluate whether those marginal improvements on ImageNet translate to improvements on similar datasets or not. To do so, we investigate twelve popular SSL frameworks on five ImageNet variants and discover that models that seem to perform well on ImageNet may experience significant performance declines on similar datasets. Specifically, state-of-the-art frameworks such as DINO and Swav, which are praised for their performance, exhibit substantial drops in performance while MoCo and Barlow Twins displays comparatively good results. As a result, we argue that otherwise good and desirable properties of models remain hidden when benchmarking is only performed on the ImageNet validation set, making us call for more adequate benchmarking. To avoid the "benchmark lottery" on ImageNet and to ensure a fair benchmarking process, we investigate the usage of a unified metric that takes into account the performance of models on other ImageNet variant datasets. Copyright © 2025, The Authors. All rights reserved.

关键词： Self-supervised learning

CORD: Co-design of Resource Allocation and Deadline Decomposition with Generative Profiling

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Gifford, Robert Eisenklam, Abby Bondar, Georgiy A. Cai, Yifan Sial, Tushar Phan, Linh Thi Xuan Halder, Abhishek Department of Computer and Information Science University of Pennsylvania United States Department of Applied Mathematics University of California Santa Cruz United States Department of Aerospace Engineering Iowa State University United States

As multicore hardware is becoming increasingly common in real-time systems, traditional scheduling techniques that assume a single worst-case execution time for a task are no longer adequate, since they ignore the impact of shared resources on execution time. When tasks execute concurrently on different cores, their execution times often vary substantially with their allocated budgets of shared resources, such as cache and memory bandwidth. Even under a specific resource allocation, the resource use pattern of a task also changes with time during a job execution. It is therefore important to consider the relationship between multicore resources and execution time in task modeling and scheduling algorithm design. In this paper, we propose a much more precise execution model for DAG-based real-time tasks that captures the time-varying resource use characteristics of a task under different budgets of shared resources. We present a generative resource profiling algorithm that efficiently predicts, from limited measurement data, the resource profile of a task at any time during its execution under a given resource budget. The generative profiles can then be used to construct the execution models for tasks, using which one can make informed resource allocation decisions. We further introduce a multicore resource allocation and deadline decomposition co-design technique for DAG-based tasks that leverages the generated execution models to jointly allocate resources and deadlines to subtasks, to maximize resource efficiency and schedulability. Our evaluation results show that our generative profiling algorithm achieves high accuracy while being efficient, and that our co-allocation technique substantially improves schedulability compared to a state-of-the-art deadline decomposition method. Copyright © 2025, The Authors. All rights reserved.

关键词： Budget control

Towards Efficient Training of Graph Neural Networks: A Multiscale Approach

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Gal, Eshed Eliasof, Moshe Schönlieb, Carola-Bibiane Haber, Eldad Treister, Eran Department of Computer Science Ben-Gurion University of the Negev Department of Applied Mathematics and Theoretical Physics University of Cambridge United Kingdom Department of Earth Ocean and Atmospheric Sciences University of British Columbia Canada

Graph Neural Networks (GNNs) have emerged as a powerful tool for learning and inferring from graph-structured data, and are widely used in a variety of applications, often considering large amounts of data and large graphs. However, training on such data requires large memory and extensive computations. In this paper, we introduce a novel framework for efficient multiscale training of GNNs, designed to integrate information across multiscale representations of a graph. Our approach leverages a hierarchical graph representation, taking advantage of coarse graph scales in the training process, where each coarse scale graph has fewer nodes and edges. Based on this approach, we propose a suite of GNN training methods: such as coarse-to-fine, sub-to-full, and multiscale gradient computation. We demonstrate the effectiveness of our methods on various datasets and learning tasks. Copyright © 2025, The Authors. All rights reserved.

关键词： Graph neural networks

An Algorithm for Illuminating n Nonoverlapping Circular Discs’ Boundaries on the Plane with Application to Tree Stem Illumination Problem

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Sukkasem, Phapaengmuang Chaidee, Supanut Khokthong, Watit Program in Applied Mathematics Department of Mathematics Faculty of Science Chiang Mai University Chiang Mai50200 Thailand Department of Mathematics Faculty of Science Chiang Mai University Thailand Forest Restoration Research Unit Department of Biology Faculty of Science Chiang Mai University Chiang Mai50200 Thailand Environmental Science Research Centre Faculty of Science Chiang Mai University Chiang Mai50200 Thailand

Given a set of n nonoverlapping circular discs on a plane, we aim to determine possible positions of points (referred to as cameras) that could fully illuminate all the circular discs’ boundaries. This work presents a geometric approach for determining feasible camera positions that would provide total illumination of all circular discs. The Laguerre Delaunay triangulation, coupled with the intersection of slabs formed by the boundaries of circular discs, is employed to form the region that satisfies the given conditions. The experiment is conducted using a set of randomly positioned circular discs on a plane. This study has the potential to address the issue of illumination in forests by utilizing a LiDAR camera to determine the possible number and placement of cameras that can effectively illuminate trees within a forest. © 2025, CC BY.

关键词： Laser recording

Forecasting day-ahead electric power prices with functional data analysis

学校读者我要写书评

暂无评论

Frontiers in Energy Research 2025年 13卷

作者： Jan, Faheem Iftikhar, Hasnain Tahir, Mehwish Khan, Mehak Department of Mathematics and Statistics Bacha Khan University Charsadda Pakistan Department of Statistics Quaid-i-Azam University Islamabad Pakistan Department of Mathematics Shaheed Benazir Bhutto Women University Peshawar Pakistan Department of Computer Science Electrical Engineering and Mathematical Sciences Western Norway University of Applied Sciences Bergen Norway

Day-ahead electricity prices in today’s competitive electric power markets have complex features such as high frequency, high volatility, non-linearity, non-stationarity, mean reversion, multiple periodicities, and calendar effects. These complicated features make price forecasting difficult. To address this, this research examines the application of functional data analysis to forecasting day-ahead electric power prices. Compared to classical time series forecasting approaches, functional data analysis is more appealing since it anticipates the daily profile, allowing for short-term projections. This technique uses a functional autoregressive ((Formula presented.) AR) and a functional autoregressive with exogenous predictors ((Formula presented.) AR (Formula presented.)) model to predict the next-day electric power prices. In addition, standard time-series forecasting models, including autoregressive (AR) AR (Formula presented.), autoregressive integrated moving average (ARIMA), and ARIMA (Formula presented.) are also utilized for comparison. The model’s prediction performance was evaluated using data on electricity prices from the British electricity market, considering forecast error indicators and the same forecast statistical test. The results show that the proposed functional models ((Formula presented.) AR and (Formula presented.) AR (Formula presented.)) outperform standard time series models. In comparison to the benchmark models (AR, AR (Formula presented.), ARIMA, ARIMA (Formula presented.), and the proposed (Formula presented.) AR model), the (Formula presented.) AR (Formula presented.) model reduces: the day-ahead forecasting average MAPE by ranges of 5.02%–45.77%, 4.07%–40.63%, 3.80%–38.99%, 1.90%–24.22%, and 0.95%–13.78%;MAE by ranges of 9.43%–69.32%, 5.17%–65.48%, 6.04%–59.16%, 3.02%–42.01%, and 1.51%–26.59%;RMSE by ranges of 8.98%–40.97%, 6.68%–34.03%, 4.22%–24.58%, 3.91%–23.20%, and 2.30%–15.11%. Furthermore, compared with the literature-proposed b

关键词： classical time series models day-ahead electricity price forecasting electric power market functional data analysis functional time series models