Uncertainty-aware semi-supervised method using large unlabeled and limited labeled COVID-19 data
The new coronavirus has caused more than one million deaths and continues to spread
rapidly. This virus targets the lungs, causing respiratory distress which can be mild or …
rapidly. This virus targets the lungs, causing respiratory distress which can be mild or …
Segmentation of intracranial hemorrhage using semi-supervised multi-task attention-based U-net
Intracranial Hemorrhage (ICH) has high rates of mortality, and risk factors associated with it
are sometimes nearly impossible to avoid. Previous techniques to detect ICH using machine …
are sometimes nearly impossible to avoid. Previous techniques to detect ICH using machine …
Semi-supervised ensemble learning for dealing with inaccurate and incomplete supervision
In real-world tasks, obtaining a large set of noise-free data can be prohibitively expensive.
Therefore, recent research tries to enable machine learning to work with weakly supervised …
Therefore, recent research tries to enable machine learning to work with weakly supervised …
Multi-platform model processing and execution management engine
RA Nendorf, N Malpekar, MV Slusar… - US Patent …, 2020 - Google Patents
Systems and methods are disclosed for managing the pro cessing and execution of models
that may have been devel oped on a variety of platforms. A multi-model execution module …
that may have been devel oped on a variety of platforms. A multi-model execution module …
Develo** and Evaluating Algorithms for Fixing Omission and Commission Errors in Structured Data
M Nashaat Ali Elmowafy - 2020 - era.library.ualberta.ca
Mona Nashaat Ali Elmowafy Page 1 Develo** and Evaluating Algorithms for Fixing Omission
and Commission Errors in Structured Data by Mona Nashaat Ali Elmowafy A thesis submitted in …
and Commission Errors in Structured Data by Mona Nashaat Ali Elmowafy A thesis submitted in …
A Robust Semi-Supervised Learning Framework for Predicting Breast Cancer Recurrence
A Erekat - 2021 - search.proquest.com
Predicting breast cancer recurrence is a substantial challenge in breast cancer
management. Current prediction tools rely only on phenotypic (non-personalized) measures …
management. Current prediction tools rely only on phenotypic (non-personalized) measures …
Performance Analysis of Semi-Supervised Learning Methods under Different Missing Label Patterns
In this study, we analyze the performance of semi-supervised learning methods under
different missing label patterns and missing label proportions. Some semi-supervised …
different missing label patterns and missing label proportions. Some semi-supervised …
Multi-platform machine learning systems
P O'reilly, N Malpekar, RA Nendorf, BT Le… - US Patent …, 2023 - Google Patents
Aspects of the disclosure relate to systems, methods, and computing devices for managing
the processing and execution of machine learning classifiers across a variety of platforms …
the processing and execution of machine learning classifiers across a variety of platforms …