Results 71 to 80 of about 111,792 (225)

MGM as a Large‐Scale Pretrained Foundation Model for Microbiome Analyses in Diverse Contexts

Advanced Science, EarlyView.
We present the Microbial General Model (MGM), a transformer‐based foundation model pretrained on over 260,000 microbiome samples. MGM learns contextualized microbial representations via self‐supervised language modeling, enabling robust transfer learning, cross‐regional generalization, keystone taxa discovery, and prompt‐guided generation of realistic,
Haohong Zhang, Yuli Zhang, Zixin Kang, Jiayun Xiong, Ronghua Yang, Kang Ning   +5 more
wiley   +1 more source

Implementasi Extra Trees Classifier dengan Optimasi Grid Search CV pada Prediksi Tingkat Adaptasi [PDF]

Innovations and advancements in technology are always happening, especially in the fields of communication technology, education technology, and information technology.
Aina, Listya Nur Aina
core  

Long Short Term Memory Hyperparameter Optimization for a Neural Network Based Emotion Recognition Framework

IEEE Access, 2018
Recently, emotion recognition using low-cost wearable sensors based on electroencephalogram and blood volume pulse has received much attention. Long short-term memory (LSTM) networks, a special type of recurrent neural networks, have been applied ...
Bahareh Nakisa, Mohammad Naim Rastgoo, Andry Rakotonirainy, Frederic Maire, Vinod Chandran   +4 more
doaj   +1 more source

Evaluating the Utilities of Foundation Models in Single‐Cell Data Analysis

Advanced Science, EarlyView.
This study delivers the first systematic, task‐level evaluation of single‐cell foundation models across eight core analytical tasks. By benchmarking 10 leading models with the scEval framework, it reveals where foundation models truly add value, where task‐specific methods still dominate, and provides concrete, reproducible guidelines to steer the next
Tianyu Liu, Kexing Li, Yuge Wang, Hongyu Li, Hongyu Zhao   +4 more
wiley   +1 more source

Random Search Hyperparameter Optimization for BPNN to Forecasting Cattle Population [PDF]

E3S Web of Conferences
Backpropagation Neural Network (BPNN) is a suitable method for predicting the future. It has weaknesses, namely poor convergence speed and instability, requiring parameter tuning to overcome speed problems, and having a high bias.
Khotimah Bain Khusnul, Agustina Fitri, Puspitarini Oktavia Rahayu, Husni, Anamisa Devie Rosa, Prayugo Natasha, Putri Aisyah Meta Sari   +6 more
doaj   +1 more source

De Novo Multi‐Mechanism Antimicrobial Peptide Design via Multimodal Deep Learning

Advanced Science, EarlyView.
Current AI‐driven peptide discovery often overlooks complex structural data. This study presents M3‐CAD, a generative pipeline that leverages 3D voxel coloring and a massive database of over 12 000 peptides to capture nuanced physicochemical contexts.
Xiaojuan Li, Haifan Gong, Yue Wang, Yinuo Zhao, Lixiang Li, Peijing Bao, Qingzhou Kong, Jialu Fu, Boyao Wan, Yumeng Zhang, Jinghui Zhang, Jiekun Ni, Zhongxue Han, Xueping Nan, Kunping Ju, Longfei Sun, Yuerui Ma, Huijun Chang, Mengqi Zheng, Yanbo Yu, Xiaoyun Yang, Xiuli Zuo, Haina Wang, Yanqing Li   +23 more
wiley   +1 more source

High‐Fidelity Synthetic Data Replicates Clinical Prediction Performance in a Million‐Patient Diabetes Cohort

Advanced Science, EarlyView.
This study generates high‐fidelity synthetic longitudinal records for a million‐patient diabetes cohort, successfully replicating clinical predictive performance. However, deeper analysis reveals algorithmic biases and trajectory inconsistencies that escape standard quality metrics. These findings challenge current validation norms, demonstrating why a
Francisco Ortuño, Víctor M. de la Oliva Roque, Javier‐Ignacio Ramirez‐Lopez, David P. Kreil, Joaquín Dopazo, Carlos Loucera   +5 more
wiley   +1 more source

Deep Learning‐Powered Scalable Cancer Organ Chip for Cancer Precision Medicine

Advanced Science, EarlyView.
This scalable, low‐cost Organ Chip platform, made via injection molding, uses capillary pinning for hydrogel confinement and supports versatile tissue coculture and robust imaging. Deep learning enables label‐free, sensitive phenotypic analysis.
Yu‐Chieh Yuan, Beibei Xu, Jenna McCormack, XuHai Huang, Jingzhe Ma, Thomas Marshall, Yacong Sun, Hardeep Singh, Alyssa Fanelli, Gauri Kulkarni, Ji Hye Seo, Paige Gilbride, Bing Wei, Bo Wang, Yanyan Liu, Fei Ma, Lin Zhou, Shuyang Wang, Xiaohua Qian, Zhiyong Xie, Polina Golland, Longlong Si, Yu Shrike Zhang, Xin Xie, Haiqing Bai   +24 more
wiley   +1 more source

Structural Eigenmodes of the Brain to Improve the Source Localization of EEG: Application to Epileptiform Activity

Advanced Science, EarlyView.
Geometry and connectivity are complementary structures, which have demonstrated their ability to represent the brain's functional activity. This study evaluates geometric and connectome eigenmodes as biologically informed constraints for EEG source localization.
Pok Him Siu, Philippa J. Karoly, Sina Mansour L., Artemio Soto‐Breceda, Levin Kuhlmann, Mark J. Cook, David B. Grayden   +6 more
wiley   +1 more source

Ensemble effort estimation with metaheuristic hyperparameters and weight optimization for achieving accuracy.

PLoS ONE
Software development effort estimation (SDEE) is recognized as vital activity for effective project management since under or over estimating can lead to unsuccessful utilization of project resources.
Anum Yasmin, Wasi Haider Butt, Ali Daud
doaj   +1 more source