線形二次レギュレータのための値反復および方策反復の収束性とロバスト性に関する拡張解析
本稿では、線形二次レギュレータ問題に対する値反復および方策反復アルゴリズムの指数関数的収束性を保証するための条件を緩和し、さらに、システム行列の不確実性に対するロバスト性を解析することで、これらのアルゴリズムのより深い理解を提供する。
Convergence and Robustness Analysis of Value and Policy Iteration for the Discrete-Time Linear Quadratic Regulator Problem
This paper investigates and extends the convergence and robustness properties of value and policy iteration algorithms for solving discrete-time linear quadratic regulator problems, both in the ideal case of known system models and in the practical scenario of model uncertainties.
Automatic Identification of Political Hate Speech in Bangla Social Media Text using Recurrent Neural Networks and Traditional Machine Learning
This research explores the use of machine learning, particularly recurrent neural networks (LSTM, GRU), along with traditional methods (SVM, NB), to automatically identify political hate speech in Bangla social media text, comparing different feature extraction techniques and highlighting the challenges posed by the understudied nature of Bangla NLP.
基於大數據技術的社群網路貼文實時壓力檢測：比較機器學習與深度學習模型
本文探討利用機器學習和深度學習模型，分析 Reddit 社群網路貼文，進行實時壓力檢測的可行性，並比較不同模型的效能。
소셜 미디어 게시물의 스트레스 분석을 위한 다양한 머신러닝 및 딥러닝 모델 비교 분석 및 실시간 스트레스 탐지 시스템 구축
본 연구는 소셜 미디어 게시물에서 스트레스를 탐지하기 위해 다양한 머신러닝 및 딥러닝 모델을 비교 분석하고, 실시간 스트레스 탐지 시스템을 구축하는 것을 목표로 한다.
Stress Detection in Reddit Posts Using Machine Learning and Big Data Technologies
This research paper presents a real-time system for detecting stress in Reddit posts using machine learning and big data technologies, achieving promising results with the XLNet model and highlighting the potential of such systems for timely user support on social media platforms.
基於串聯擴散橋樑模型的時間序列預測
本文提出了一種基於串聯擴散橋樑模型（S2DBM）的新型時間序列預測方法，該方法利用布朗橋過程來減少擴散估計中的隨機性，並通過結合歷史時間序列數據中的先驗信息和條件來提高預測準確性。
시계열 데이터 예측을 위한 시계열 간 확산 브리지 모델
본 논문에서는 시계열 데이터의 포인트-투-포인트 예측 정확도를 향상시키기 위해 브라운 브리지 확산 프로세스를 활용한 새로운 확산 기반 시계열 예측 모델인 S2DBM(Series-to-Series Diffusion Bridge Model)을 제안합니다.
疑似ラベルの領域分割に基づく半教師あり医用画像セグメンテーションのためのシナジーガイド型学習手法
本稿では、ラベル付けされたデータが少ない状況下での医用画像セグメンテーションの精度向上のため、疑似ラベルの信頼性評価に基づいて学習領域を分割し、それぞれに適した損失関数を適用するSGRS-Netと呼ばれる新たな半教師あり学習フレームワークを提案する。
Series-to-Series Diffusion Bridge Model for Improved Time Series Forecasting
This paper introduces S2DBM, a novel diffusion-based time series forecasting model that leverages the Brownian Bridge process and linear conditioning methods to enhance the accuracy and stability of both point-to-point and probabilistic predictions.
© 2024 by Linnk AI