Feature abstraction for early detection of multi-type of dementia with sparse auto-encoder

Alkabawi, Elham M.; Hilal, Allaa R.; Basir, Otman

doi:10.1109/smc.2017.8123168

Cited by 10 publications

(8 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, we observe that in the Loan dataset the accuracy goes up at some point while increasing the NCP score, before going back down. This may be an indication that in this dataset, for some features, generalizing the feature to a smaller domain actually has a positive effect on the model's accuracy, as suggested by previous works on feature abstraction [2].…”

Section: Discussion Of Resultssupporting

confidence: 61%

“…This principle, known as data minimization, requires that organizations and governments collect only data that is needed to achieve the purpose at hand. The California Privacy Rights Act (CPRA) 2 , that supersedes the California Consumer Privacy Act (CCPA) and will be fully operative in 2023, also includes the data minimization principle. Organizations are expected to be able to demonstrate that the data they collect is absolutely necessary, by showing concrete measures that were taken to minimize the amount of data used for a given purpose.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Data minimization for GDPR compliance in machine learning models

et al. 2021

View full text Add to dashboard Cite

The EU General Data Protection Regulation (GDPR) and the California Privacy Rights Act (CPRA) mandate the principle of data minimization, which requires that only data necessary to fulfill a certain purpose be collected. However, it can often be difficult to determine the minimal amount of data required, especially in complex machine learning models such as deep neural networks. We present a first-of-a-kind method to reduce the amount of personal data needed to perform predictions with a machine learning model, by removing or generalizing some of the input features of the runtime data. Our method makes use of the knowledge encoded within the model to produce a generalization that has little to no impact on its accuracy, based on knowledge distillation approaches. We show that, in some cases, less data may be collected while preserving the exact same level of model accuracy as before, and if a small deviation in accuracy is allowed, even more generalizations of the input features may be performed. We also demonstrate that when collecting the features dynamically, the generalizations may be even further improved. This method enables organizations to truly minimize the amount of data collected, thus fulfilling the data minimization requirement set out in the regulations.

show abstract

Section: Discussion Of Resultssupporting

confidence: 61%

Section: Introductionmentioning

confidence: 99%

Data minimization for GDPR compliance in machine learning models

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Three classification methods comprising elastic-net logistic regression (ENLR), multiple discriminant analysis (MDA), and SVM were used to classify AD, asymptomatic controls (CTR) and FTD participants. In [88], MRI of multi-type dementia is used to reduce the dimensionality of an extract key features using sparse autoencoder (SAE), and PCA approach used to train a LDA and LR classifiers to compare their prediction accuracy.…”

Section: A Conventional Machine Learning Approaches For Dementia Diamentioning

confidence: 99%

“…Figure 7 shows a deep learning based biomarker using Softmax as a classifier for early diagnosis of dementia. Figure 7: Deep learning CNN based hierarchical feature learning and classification to prognosis dementia using Softmax classifier [88], [92], [93].…”

Section: B Deep Learning Approaches For Dementia Diagnosismentioning

confidence: 99%

“…With the comparison of SVM (67.01%), Logistics regression(67.04%) and CNN with softmax layer (61%), the proposed method in [88], [93], combined convolutional neural networks and logistic regression show better performance (74.93%) by training them on MRI scans of the AD. Where they used six convolutional layers, six max-pooling layers, six local normalization layers, three fully connection layers, and two dropout layers in CNN learning.…”

Section: B Deep Learning Approaches For Dementia Diagnosismentioning

confidence: 99%

See 1 more Smart Citation

Neuroimaging and Machine Learning for Dementia Diagnosis: Recent Advancements and Future Prospects

Ahmed

Zhang

Feng

et al. 2019

IEEE Rev. Biomed. Eng.

101

View full text Add to dashboard Cite

Dementia, a chronic and progressive cognitive declination of brain function due to disease or impairment, is becoming more prevalent due to the aging population. A major challenge in dementia is achieving the accurate and timely diagnosis. In recent years, neuroimaging with computer-aided algorithms have made remarkable advances in addressing this challenge. Much of the success of these approaches can be attributed to the application of machine learning and deep learning techniques for neuroimaging. In this review paper, we present a comprehensive survey of automated diagnostic approaches for dementia using medical image analysis and machine learning algorithms published in the previous five years (2013-2018). Based on the rigorous review of the existing works, we have found that while most of the studies focused on Alzheimer's disease have demonstrated reasonable performance, the identification of other types of dementia remains a major challenge. Hybrid imaging analysis through deep learning approaches may hold promises for early diagnosis of these other types of dementia, and warrant further investigation in the future. The main contributions of this review paper are as follows: 1. Based on the detailed analysis of the existing literature, this paper discusses the most recent neuroimaging procedures for dementia diagnosis, and 2. It systematically explains the machine learning techniques and, in particular, deep learning approaches for early detection of dementia.

show abstract

Early-Stage Dementia Detection by Optimize Feature Weights with Ensemble Learning

Mahajan

Srivastava

2023

Communications in Computer and Information Science

View full text Add to dashboard Cite

Feature abstraction for early detection of multi-type of dementia with sparse auto-encoder

Cited by 10 publications

References 13 publications

Data minimization for GDPR compliance in machine learning models

Data minimization for GDPR compliance in machine learning models

Neuroimaging and Machine Learning for Dementia Diagnosis: Recent Advancements and Future Prospects

Early-Stage Dementia Detection by Optimize Feature Weights with Ensemble Learning

Contact Info

Product

Resources

About