Publications by authors named "Ziad Akram Ali Hammouri"
Physical activity is deemed critical to successful ageing. Despite evidence and progress, there is still a need to determine more precisely the direction, magnitude, intensity, and volume of physical activity that should be performed on a daily basis to effectively promote the health of individuals. This study aimed to assess the clinical validity of new physical activity phenotypes derived from a novel distributional functional analysis of accelerometer data in older adults.
View Article and Find Full Text PDFFast Support Vector Classification for Large-Scale Problems.
IEEE Trans Pattern Anal Mach Intell
October 2022
Article Synopsis
- - The fast support vector classifier (FSVC) is an innovative solution to the slow performance and high memory consumption issues of traditional support vector machine (SVM) algorithms, particularly on large datasets.
- - FSVC features include an efficient training process without numerical iterations, a reduced number of class prototypes to minimize memory usage, and a quick method for selecting kernel parameters directly from the data.
- - Compared to existing methods like Liblinear and Libsvm, FSVC is significantly faster and requires less memory while providing better performance, making it ideal for classifying large datasets on computers with limited resources.