Diversity in Machine Learning

Gong, Zhiqiang; Zhong, Ping; Hu, Weidong

doi:10.1109/access.2019.2917620

Cited by 163 publications

(77 citation statements)

References 96 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…ML tasks are classified into four broad categories, namely supervised learning, unsupervised learning, active learning and reinforcement learning [ 11 – 13 ]. Supervised learning infers a function from the labeled training data, unsupervised learning infers a function from unlabeled training data, and active learning infers a function by choosing the most informative sample for labeling to train the model [ 12 ], while reinforcement learning interacts with a dynamic environment [ 10 , 12 ]. The flowchart of the training process of ML tasks including supervised learning, unsupervised learning and active learning is shown in Fig.…”

Section: Introductionmentioning

confidence: 99%

Predictive Supervised Machine Learning Models for Diabetes Mellitus

2020

View full text Add to dashboard Cite

Diabetes mellitus (DM) is one of the deadliest diseases in the world, especially in developed nations. In recent years, it has become rampant in the developing nations such as Nigeria, posing more threats to individuals in the latter than those in the former. More than 415 million people were reported to suffer from DM worldwide as of 2015, with type 2 of the disease accounting for approximately 90% of the cases. The number of people with DM is expected to rise to 592 million by the year 2035. Therefore, DM is one of the growing public health concerns in Nigeria. In this study, the diagnostic dataset of DM type 2 was collected from the Murtala Mohammed Specialist Hospital, Kano, and used to develop predictive supervised machine learning models based on logistic regression, support vector machine, K-nearest neighbor, random forest, naive Bayes and gradient booting algorithms. The random forest predictive learning-based model appeared to be one of the best developed models with 88.76% in terms of accuracy; however, in terms of receiver operating characteristic curve, random forest and gradient booting predictive learning-based models were found to be the best predictive learning models with 86.28% predictive ability, respectively. Keywords Machine learning • Predictive model • Diabetes mellitus • Diabetes mellitus type 2 • Random forest This article is part of the topical collection "Advances in Computational Approaches for Artificial Intelligence, Image Processing, IoT and Cloud Applications" guest edited by Bhanu Prakash K N and M. Shivakumar".

show abstract

Section: Introductionmentioning

confidence: 99%

Predictive Supervised Machine Learning Models for Diabetes Mellitus

2020

View full text Add to dashboard Cite

show abstract

“…Second, there are fewer training examples of severely degenerated discs, and many degenerative phenotypes exist. Healthy discs are often surrounded by normal presenting anatomy, while severe IVDD is associated with fattier vertebral bone marrow, narrowed spinal canal, and even signal voids due to the vacuum phenomena . Finally, Dice coefficient and percent volume difference are sensitive to segmented tissue size, and given the smaller disc volume in severe IVDD, single pixel errors disproportionately affected these results.…”

Section: Discussionmentioning

confidence: 99%

Lumbar intervertebral disc characterization through quantitative MRI analysis: An automatic voxel‐based relaxometry approach

Iriondo

Pedoia

Majumdar

2020

Magnetic Resonance in Med

View full text Add to dashboard Cite

Purpose:To develop an automated pipeline based on convolutional neural networks to segment lumbar intervertebral discs and characterize their biochemical composition using voxel-based relaxometry, and establish local associations with clinical measures of disability, muscle changes, and other symptoms of lower back pain. Methods: This work proposes a new methodology using MRI (n = 31, across the spectrum of disc degeneration) that combines deep learning-based segmentation, atlas-based registration, and statistical parametric mapping for voxel-based analysis of T 1ρ and T 2 relaxation time maps to characterize disc degeneration and its associated disability. Results: Across degenerative grades, the segmentation algorithm produced accurate, high-confidence segmentations of the lumbar discs in two independent data sets.Manually and automatically extracted mean disc T 1ρ and T 2 relaxation times were in high agreement for all discs with minimal bias. On a voxel-by-voxel basis, imagingbased degenerative grades were strongly negatively correlated with T 1ρ and T 2 , particularly in the nucleus. Stratifying patients by disability grades revealed significant differences in the relaxation maps between minimal/moderate versus severe disability: The average T 1ρ relaxation maps from the minimal/moderate disability group showed clear annulus nucleus distinction with a visible midline, whereas the severe disability group had lower average T 1ρ values with a homogeneous distribution. Conclusion: This work presented a scalable pipeline for fast, automated assessment of disc relaxation times, and voxel-based relaxometry that overcomes limitations of current region of interest-based analysis methods and may enable greater insights and associations between disc degeneration, disability, and lower back pain. K E Y W O R D Sdeep learning, registration, relaxometry, segmentation, spine

show abstract

“…Note that the diversity described in this study involves the creation of two data sets either using a pair of transmitter/receiver arrangement or a single transmitter and two receivers. Even though this approach is similar to model diversification in machine learning , the primary use of the two sets described here is for the enhancement of the overall performance of the machine vision system as seen in Figure . Diversity implemented in this study is similar to the diversity in sensor networks and wireless systems resulting in lower error rates.…”

Section: Discussionmentioning

confidence: 99%

Pedagogy of diversity and data analytics: Theory to practice

Shankar

2019

Comp Applic In Engineering

View full text Add to dashboard Cite

A course in probability is a requirement in Baccalaureate Programs in Electrical and Computer Engineering. Students view this course as conceptual with little connection to real world problems. Efforts have been undertaken to mitigate this issue and connect the conceptual topics to data analytics to make the course relevant. Data analytics based exercises are now integral to the course. Several demos were created to link statistical concepts with practice through analysis of data. In this study, a demo created to illustrate the concept of diversity to improve the performance of a machine vision system is described. It incorporates concepts of Bayes' rule, single and multiple random variables, goodness fit tests, random number simulation and data analytics to illustrate the pedagogy of diversity and associated data processing. Student survey results suggest that these demos enhance that the learning experience in the engineering probability course. K E Y W O R D Sdiversity, engineering probability, Matlab, receiver operating characteristics curves

show abstract

Diversity in Machine Learning

Cited by 163 publications

References 96 publications

Predictive Supervised Machine Learning Models for Diabetes Mellitus

Predictive Supervised Machine Learning Models for Diabetes Mellitus

Lumbar intervertebral disc characterization through quantitative MRI analysis: An automatic voxel‐based relaxometry approach

Pedagogy of diversity and data analytics: Theory to practice

Contact Info

Product

Resources

About