Tanat Tonguthaisri, CISSP?的动态

?? Exciting news! Just released a new blog post on an innovative approach to Data-Free Knowledge Distillation (DFKD) methods, addressing the discrepancy between teacher's training data and real-world scenarios. Our latest work introduces AuG-KD, a method that effectively aligns student-domain data with the teacher domain, ensuring superior performance in real-world applications. Check out the full post on ArXiv at: https://bit.ly/4c7kGj3 #research #datascience #machinelearning #knowledge distillation

We recently examined the composition of MATs in terms of school phase, and uncovered the question, why are there so few secondary-only MATs? ?? Our new DataLab blog unpacks this topic by exploring why this may be and whether this may change in the future. Read the key findings: https://buff.ly/4auKbJh #MATs #MAT #Education

Check out how the #PISDMTSS universal screeners & data tracker are being used (Turner ES) to identify students in need of Tier 3 intensive intervention & determine the needed interventions! Thank you, Ms. Caplan, for your support of our #PISDMathChat mathematicians! #RTIaW #AtPLC

Excited to present Vision-LSTM (ViL) at two ICML workshops (NGSM and LSFM). We also updated the arxiv paper with: ?? Better training pipeline/models ?? Pre-training and inference FLOPS ?? Evaluation on more datasets (segmentation, transfer classification) ?? https://lnkd.in/dWQjB28Y

How can the use of item response theory improve the ways we measure, describe and develop teachers' knowledge of effective use of technology? Our latest paper explores how Rasch modeling can be used to provide substantive qualitative explanations of quantitative measurement scores derived from TPACK survey tools. https://lnkd.in/gf_rPBiM #TPACK #educationtechnology #PCK SITE Conference - Society for Information Technology & Teacher Education #edtech #teachereducation

I'm happy to share our latest research paper is now available in MDPI's Machine Learning and Knowledge Extraction (MAKE MDPI) journal. The paper, "Ensemble learning with highly variable class-based performance" proposes a novel model-agnostic approach to leveraging the class-dependent decisions of base classifiers in a machine learning ensemble. Our approach outperforms the state-of-the-art class-specific soft voting (CSSV) methodology on 10 open-source UCI multi-class datasets. Edward Ratner Kallin Carolus Khan https://lnkd.in/gBagx_Ns #ensemblelearning #MLensembles #ExtremeLearningMachines #ELM #machinelearning #artificialintelligence

Struggling with the Data Insights section of the #GMAT? Discover the ultimate timing strategy in this article! Learn to master each question type, manage your time effectively, and boost your efficiency. Don't miss out on these essential tips to ace the DI section and maximize your GMAT score! ?? Read the article here: https://ow.ly/nyYm50StJU9 #GMATFocus #GMATPrep #GMATTips #GMATStrategies #TargetTestPrep #TTP #PrepareWithTheBestRockTheTest

Creating great instruction is part art, part science. For the science part, it can help to have some guidelines to steer you. We use learning standards here at BCL for that purpose. Feel free to check them out and use or adapt them in any way you wish. Here's a link: https://lnkd.in/eHF3WEHT

?? Excited to share a comprehensive review on Few-shot Learning on Heterogeneous Graphs (FLHG) in our latest blog post. Discover the challenges, progress, and promising prospects in this rapidly evolving field. This paper is the first of its kind, offering a deep dive into FLHG methods and applications. Check it out here: https://bit.ly/3x6rD3G #MachineLearning #GraphTheory #FLHG #ResearchUpdate

Most read papers of IJPEDS: