Algorithms MDPI’s Post

📢 Excited to share our latest blog post on "Improving Label Error Detection and Elimination with Uncertainty Quantification"! Identifying and handling label errors is crucial for enhancing the accuracy of supervised machine learning models. Our paper presents novel, model-agnostic algorithms for Uncertainty Quantification-Based Label Error Detection (UQ-LED), surpassing state-of-the-art confident learning in identifying label errors. Find out more about our groundbreaking approach and its impact on dataset cleaning and model accuracy here: https://bit.ly/44SK34U #MachineLearning #DataScience #UncertaintyQuantification

Delighted to share that two of our papers have been accepted recently: 1️⃣ "No Dimensional Sampling Coresets for Classification" by Meysam Alishahi and Jeff M. Phillips, accepted at the International Conference on Machine Learning (ICML), July 2024. 2️⃣ "Linear Distance Metric Learning with Noisy Labels" by Meysam Alishahi, Anna Little, and Jeff M. Phillips, accepted in the Journal of Machine Learning Research (JMLR), April 2024. #MachineLearning #ICML #JMLR #Research #MetricLearning #Coresets

Could Artificial Intelligence be appropriate to a degrowth context? A conviviality perspective on Machine Learning. My paper is out in the journal GAIA - Ecological Perspectives for Science and Society! In this paper, I apply the Matrix of Convivial Technology designed by Vetter to a particular application of machine learning (predictive maintenance), and find three key limitations to its conviviality: (1) the high complexity of machine learning, (2) the environmental impacts of machine learning, and (3) the (size of the) infrastructure it relies on. Make sure to check out the full article here https://lnkd.in/dt_QKmDw Thank you to everyone at GAIA for their efforts!

Registration is filling up! Drs Niklas Korsbo and Lorenzo Contento are bringing an in-depth, hands-on workshop to PAGE 2024. Those who attend will gain practical skills in: ✔️ hybrid mechanistic and machine learning modeling (NeuralODE, UDE, and DeepNLME), ✔️ a comprehension of how complex covariates can help forecast longitudinal outcomes, and ✔️ insights into the possible implications of this technology. Full abstract and registration details are here: https://loom.ly/_aGYGb0 #PAGE2024 #neuralnetworks #machinelearning #scientificmodeling #healthcaredata

Faced with the dilemma of selecting either scientific modeling or machine learning, our scientists opted to merge the two disciplines instead! With DeepPumas, scientists can leverage scientific laws governing the system and machine learning to bring the best knowledge to support future decisions. Watch this quick video for details: https://loom.ly/T7LuVUA Or, sign up for an upcoming workshop: https://loom.ly/_aGYGb0 #PAGE2024 #neuralnetworks #machinelearning #scientificmodeling #healthcaredata

 🔥 Read our Highly Cited Paper  📚 Impacts of Feature Selection on Predicting Machine Failures by Machine Learning Algorithms 🔗 https://lnkd.in/gzfDvB-7 👨🔬 by Francisco Elânio Bezerra et al. 🏫 USP - Universidade de São Paulo #machinelearning #machinefailure

GEOMETRY-INFORMED MACHINE LEARNING IN BRAIN Bridging tuning and invariance with equivariant neuronal representations "We propose that equivariance is a prevalent computation of populations of biological neurons to gradually achieve invariance through structured tuning." More information on GEOMETRY-INFORMED MACHINE LEARNING https://lnkd.in/eGAtMgJW

Excited to share our latest achievement! Our research paper on "Traffic Prediction Analysis Using Machine Learning Methodologies" has been published in IEEE. 📚🔍 Proud to contribute to advancements in transportation systems and leverage machine learning for more efficient traffic predictions. Grateful for the collaborative efforts of our team. Read the full paper here [https://lnkd.in/gdyAm2-g]. #Research #MachineLearning #TrafficPrediction #IEEEpublication

Although the process was lengthy, this work has finally been published and is available open-access at the Machine Learning: Science & Technology journal. Application of #machinelearning model (semantic segmentation + classifier) on 2D images of simulated CRES electrons to increase statistical sensitivity to #neutrinomass measurement. In this study, we show a relative gain in efficiency of 24.1% over traditional image reconstruction techniques. Thanks to the ML team at Project 8 and Yale University for the support! 🙏 https://lnkd.in/eWh-R9tk