Frontiers in Artificial Intelligence最新文献

{"title":"Advancements in cache management: a review of machine learning innovations for enhanced performance and security.","authors":"Keshav Krishna","doi":"10.3389/frai.2025.1441250","DOIUrl":"https://doi.org/10.3389/frai.2025.1441250","url":null,"abstract":"<p><p>Machine learning techniques have emerged as a promising tool for efficient cache management, helping optimize cache performance and fortify against security threats. The range of machine learning is vast, from reinforcement learning-based cache replacement policies to Long Short-Term Memory (LSTM) models predicting content characteristics for caching decisions. Diverse techniques such as imitation learning, reinforcement learning, and neural networks are extensively useful in cache-based attack detection, dynamic cache management, and content caching in edge networks. The versatility of machine learning techniques enables them to tackle various cache management challenges, from adapting to workload characteristics to improving cache hit rates in content delivery networks. A comprehensive review of various machine learning approaches for cache management is presented, which helps the community learn how machine learning is used to solve practical challenges in cache management. It includes reinforcement learning, deep learning, and imitation learning-driven cache replacement in hardware caches. Information on content caching strategies and dynamic cache management using various machine learning techniques in cloud and edge computing environments is also presented. Machine learning-driven methods to mitigate security threats in cache management have also been discussed.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1441250"},"PeriodicalIF":3.0,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11893820/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143606598","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transfer learning for predicting of gross domestic product growth based on remittance inflows using RNN-LSTM hybrid model: a case study of The Gambia.","authors":"Haruna Jallow, Ronald Waweru Mwangi, Alieu Gibba, Herbert Imboga","doi":"10.3389/frai.2025.1510341","DOIUrl":"10.3389/frai.2025.1510341","url":null,"abstract":"<p><p>Insights into the magnitude and performance of an economy are crucial, with the growth rate of real GDP frequently used as a key indicator of economic health, highlighting the importance of the Gross Domestic Product (GDP). Additionally, remittances have drawn considerable global interest in recent years, particularly in The Gambia. This study introduces an innovative model, a hybrid of recurrent neural network and long-short-term memory (RNN-LSTM), to predict GDP growth based on remittance inflows in The Gambia. The model integrates data sourced both from the World Bank Development Indicators and the Central Bank of The Gambia (1966-2022). Pearson's correlation was applied to detect and choose the variables that exhibit the strongest relationship with GDP and remittances. Furthermore, a parameter transfer learning technique was employed to enhance the model's predictive accuracy. The hyperparameters of the model were fine-tuned through a random search process, and its effectiveness was assessed using RMSE, MAE, MAPE, and R² metrics. The research results show, first, that it has good generalization capacity, with stable applicability in predicting GDP growth based on remittance inflows. Second, as compared to standalone models the suggested model surpassed in term of predicting accuracy attained the highest R² score of 91.285%. Third, the predicted outcomes further demonstrated a strong and positive relationship between remittances and short-term economic growth. This paper addresses a critical research gap by employing artificial intelligence (AI) techniques to forecast GDP based on remittance inflows.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1510341"},"PeriodicalIF":3.0,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11891165/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143597337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"General SIR model for visible and hidden epidemic dynamics.","authors":"Igor Nesteruk","doi":"10.3389/frai.2025.1559880","DOIUrl":"10.3389/frai.2025.1559880","url":null,"abstract":"<p><p>To simulate hidden epidemic dynamics connected with asymptomatic and unregistered patients, a new general SIR model was proposed. For some cases, the analytical solutions of the set of 5 differential equations were found, which allow simplifying the parameter identification procedure. Two waves of the pertussis epidemic in England in 2023 and 2024 were simulated with the assumption of zero hidden cases. The accumulated and daily numbers of cases and the duration of the second wave were predicted with rather high accuracy. If the trend will not change, the monthly figure of 9 new pertussis cases (as it was in January-February 2023) can be achieved only in May 2025. The proposed approach can be recommended for both simulations and predictions of different epidemics.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1559880"},"PeriodicalIF":3.0,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11891346/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143597108","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI, universal basic income, and power: symbolic violence in the tech elite's narrative.","authors":"Jean-Christophe Bélisle-Pipon","doi":"10.3389/frai.2025.1488457","DOIUrl":"10.3389/frai.2025.1488457","url":null,"abstract":"","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1488457"},"PeriodicalIF":3.0,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11891208/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143596864","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel approach to Indian bird species identification: employing visual-acoustic fusion techniques for improved classification accuracy.","authors":"Pralhad Gavali, J Saira Banu","doi":"10.3389/frai.2025.1527299","DOIUrl":"10.3389/frai.2025.1527299","url":null,"abstract":"<p><p>Accurate identification of bird species is essential for monitoring biodiversity, analyzing ecological patterns, assessing population health, and guiding conservation efforts. Birds serve as vital indicators of environmental change, making species identification critical for habitat protection and understanding ecosystem dynamics. With over 1,300 species, India's avifauna presents significant challenges due to morphological and acoustic similarities among species. For bird monitoring, recent work often uses acoustic sensors to collect bird sounds and an automated bird classification system to recognize bird species. Traditional machine learning requires manual feature extraction and model training to build an automated bird classification system. Automatically extracting features is now possible due to recent advances in deep learning models. This study presents a novel approach utilizing visual-acoustic fusion techniques to enhance species identification accuracy. We employ a Deep Convolutional Neural Network (DCNN) to extract features from bird images and a Long Short-Term Memory (LSTM) network to analyze bird calls. By integrating these modalities early in the classification process, our method significantly improves performance compared to traditional methods that rely on either data type alone or utilize late fusion strategies. Testing on the iBC53 (Indian Bird Call) dataset demonstrates an impressive accuracy of 94%, highlighting the effectiveness of our multi-modal fusion approach.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1527299"},"PeriodicalIF":3.0,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11885287/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143587413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting therapy dropout in chronic pain management: a machine learning approach to cannabis treatment.","authors":"Anna Visibelli, Rebecca Finetti, Bianca Roncaglia, Paolo Poli, Ottavia Spiga, Annalisa Santucci","doi":"10.3389/frai.2025.1557894","DOIUrl":"10.3389/frai.2025.1557894","url":null,"abstract":"<p><strong>Introduction: </strong>Chronic pain affects approximately 30% of the global population, posing a significant public health challenge. Despite their widespread use, traditional pharmacological treatments, such as opioids and NSAIDs, often fail to deliver adequate, long-term relief while exposing patients to risks of addiction and adverse side effects. Given these limitations, medical cannabis has emerged as a promising therapeutic alternative with both analgesic and anti-inflammatory properties. However, its clinical efficacy is hindered by high interindividual variability in treatment response and elevated dropout rates.</p><p><strong>Methods: </strong>A comprehensive dataset integrating genetic, clinical, and pharmacological information was compiled from 542 Caucasian patients undergoing cannabis-based treatment for chronic pain. A machine learning (ML) model was developed and validated to predict therapy dropout. To identify the most influential factors driving dropout, SHapley Additive exPlanations (SHAP) analysis was performed.</p><p><strong>Results: </strong>The random forest classifier demonstrated robust performance, achieving a mean accuracy of 80% and a maximum of 86%, with an AUC of 0.86. SHAP analysis revealed that high final VAS scores and elevated THC dosages were the most significant predictors of dropout, both strongly correlated with an increased likelihood of discontinuation. In contrast, baseline therapeutic benefits, CBD dosages, and the CC genotype of the rs1049353 polymorphism in the CNR1 gene were associated with improved adherence.</p><p><strong>Discussion: </strong>Our findings highlight the potential of ML and pharmacogenetics to personalize cannabis-based therapies, improving adherence and enabling more precise management of chronic pain. This research paves the way for the development of tailored therapeutic strategies that maximize the benefits of medical cannabis while minimizing its side effects.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1557894"},"PeriodicalIF":3.0,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11882547/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143574090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A hype-adjusted probability measure for NLP stock return forecasting.","authors":"Zheng Cao, Helyette Geman","doi":"10.3389/frai.2025.1527180","DOIUrl":"10.3389/frai.2025.1527180","url":null,"abstract":"<p><p>This article introduces a Hype-Adjusted Probability Measure in the context of a new Natural Language Processing (NLP) approach for stock return and volatility forecasting. A novel sentiment score equation is proposed to represent the impact of intraday news on forecasting next-period stock return and volatility for selected U.S. semiconductor tickers, a very vibrant industry sector. This work improves the forecast accuracy by addressing news bias, memory, and weight, and incorporating shifts in sentiment direction. More importantly, it extends the use of the remarkable tool of change of Probability Measure developed in the finance of Asset Pricing to NLP forecasting by constructing a Hype-Adjusted Probability Measure, obtained from a redistribution of the weights in the probability space, meant to correct for excessive or insufficient news.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1527180"},"PeriodicalIF":3.0,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11879933/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143567597","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"XAI-MRI: an ensemble dual-modality approach for 3D brain tumor segmentation using magnetic resonance imaging.","authors":"Ahmeed Suliman Farhan, Muhammad Khalid, Umar Manzoor","doi":"10.3389/frai.2025.1525240","DOIUrl":"10.3389/frai.2025.1525240","url":null,"abstract":"<p><p>Brain tumor segmentation from Magnetic Resonance Images (MRI) presents significant challenges due to the complex nature of brain tumor tissues. This complexity poses a significant challenge in distinguishing tumor tissues from healthy tissues, particularly when radiologists rely on manual segmentation. Reliable and accurate segmentation is crucial for effective tumor grading and treatment planning. In this paper, we proposed a novel ensemble dual-modality approach for 3D brain tumor segmentation using MRI. Initially, individual U-Net models are trained and evaluated on single MRI modalities (T1, T2, T1ce, and FLAIR) to establish each modality's performance. Subsequently, we trained U-net models using combinations of the best-performing modalities to exploit the complementary information and improve segmentation accuracy. Finally, we introduced the ensemble dual-modality by combining the two best-performing pre-trained dual-modalities models to enhance segmentation performance. Experimental results show that the proposed model enhanced the segmentation result and achieved a Dice Coefficient of 97.73% and a Mean IoU of 60.08%. The results illustrate that the ensemble dual-modality approach outperforms single-modality and dual-modality models. Grad-CAM visualizations are implemented, generating heat maps that highlight tumor regions and provide useful information to clinicians about how the model made the decision, increasing their confidence in using deep learning-based systems. Our code publicly available at: https://github.com/Ahmeed-Suliman-Farhan/Ensemble-Dual-Modality-Approach.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1525240"},"PeriodicalIF":3.0,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11880613/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143567604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Environment sustainability with smart grid sensor.","authors":"Sheetal Mahadik, Madhuri Gedam, Deven Shah","doi":"10.3389/frai.2024.1510410","DOIUrl":"10.3389/frai.2024.1510410","url":null,"abstract":"<p><p>Environmental sustainability is a pressing global concern, with energy conservation and efficient utilization playing a key role in its achievement. Smart grid technology has emerged as a promising solution, facilitating energy efficiency, promoting renewable energy integration, and fostering consumer engagement. But the addition of intelligent sensors to these grids has the potential to greatly increase the level of sustainability initiatives. <i>This paper highlights the role of smart grid sensors in addressing challenges like energy losses, demand-response limitations, and renewable energy integration. It explains how these sensors enable real-time monitoring, fault detection, and optimal load management to improve grid performance and reduce environmental impact.</i> This also study looks at how AI with smart grid sensor can perform real-time data monitoring, optimal energy distribution, and proactive decision support from smart grid sensors might improve environmental sustainability. <i>Furthermore, it examines advancements in sensor technologies in India, including pilot projects like the BESCOM initiative in Bangalore and Tata Power-DDL's renewable energy trading in Delhi, to showcase their practical applications and outcomes.</i> Smart sensors provide accurate tracking of energy usage trends, enhance load distribution, and advance the sensible application of renewable energy resources. These sensors aid in cutting down on energy waste and carbon emissions by interacting with customers and enabling demand-response systems. This study addresses the critical role of smart sensors in overcoming the shortcomings of conventional grids and guaranteeing a more resilient, efficient, and sustainable energy future through an extensive analysis of the literature. Grid-enabled systems, such as electric water heaters with sensor, can achieve energy savings of up to 29%. The integration of renewable energy sources through sensors enhances system efficiency, reduces reliance on fossil fuels, and optimizes supply and demand. Utilizing Internet of Things (IoT) technology enables precise monitoring of air quality, water consumption, and resource management, significantly improving environmental oversight. This integration can lead to a reduction in greenhouse gas emissions by up to 20% and water usage by 30%. <i>Lastly, the paper discusses how integrating artificial intelligence with smart grid sensors can enhance predictive maintenance, energy management, and cybersecurity, further strengthening the case for their deployment.</i></p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"7 ","pages":"1510410"},"PeriodicalIF":3.0,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11879990/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143567677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reconstructing signal during brain stimulation with Stim-BERT: a self-supervised learning model trained on millions of iEEG files.","authors":"Karthik Menon, Thomas Tcheng, Cairn Seale, David Greene, Martha Morrell, Sharanya Arcot Desai","doi":"10.3389/frai.2025.1502504","DOIUrl":"10.3389/frai.2025.1502504","url":null,"abstract":"<p><p>Brain stimulation has become a widely accepted treatment for neurological disorders such as epilepsy and Parkinson's disease. These devices not only deliver therapeutic stimulation but also record brain activity, offering valuable insights into neural dynamics. However, brain recordings during stimulation are often blanked or contaminated by artifact, posing significant challenges for analyzing the acute effects of stimulation. To address these challenges, we propose a transformer-based model, Stim-BERT, trained on a large intracranial EEG (iEEG) dataset to reconstruct brain activity lost during stimulation blanking. To train the Stim-BERT model, 4,653,720 iEEG channels from 380 RNS system patients were tokenized into 3 (or 4) frequency band bins using 1 s non-overlapping windows resulting in a total vocabulary size of 1,000 (or 10,000). Stim-BERT leverages self-supervised learning with masked tokens, inspired by BERT's success in natural language processing, and shows significant improvements over traditional interpolation methods, especially for longer blanking periods. These findings highlight the potential of transformer models for filling in missing time-series neural data, advancing neural signal processing and our efforts to understand the acute effects of brain stimulation.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1502504"},"PeriodicalIF":3.0,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11876146/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143558273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}