Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Wald-Wolfowitz Runs Test II

Wald-Wolfowitz Runs Test II

The Wald-Wolfowitz runs test, commonly referred to as the runs test, is a nonparametric test used to assess the randomness of ordered data. The test evaluates the number of runs, which are consecutive sequences of similar elements within the data. If the number of runs is significantly higher or lower than expected, the data is considered non-random, indicating a detectable pattern or structure.
For binary data, runs are identified using symbols such as + and −, or equivalently, 1s and...

Censoring Survival Data

Censoring Survival Data

Survival analysis is a statistical method used to analyze time-to-event data, often employed in fields such as medicine, engineering, and social sciences. One of the key challenges in survival analysis is dealing with incomplete data, a phenomenon known as "censoring." Censoring occurs when the event of interest (such as death, relapse, or system failure) has not occurred for some individuals by the end of the study period or is otherwise unobservable, and it might have many different...

Strategies for Assessing and Addressing Confounding

Strategies for Assessing and Addressing Confounding

Confounding is a critical issue in epidemiological studies, often leading to misleading conclusions about associations between exposures and outcomes. It occurs when the relationship between the exposure and the outcome is mixed with the effects of other factors that influence the outcome. Given that, addressing confounding is of high importance for drawing accurate inferences in research.
Confounding can be addressed at both the design phase of a study and through analytical methods after data...

Blinding

Blinding

Blinding is a commonly used method of not telling participants which treatment a subject is receiving. Blinding is a critical part of a randomized control trial or RCT. It reduces the bias that affects the results. In an RCT, blinding is used in the form of a placebo. A placebo effect occurs when untreated subjects falsely believe they have received the treatment and report improved symptoms. A placebo or a dummy treatment is administered to subjects to negate the bias caused by such an effect.

Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches

Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches

Biopharmaceutical studies constitute a vital field aiming to enhance drug delivery methods and refine therapeutic approaches, drawing upon diverse interdisciplinary knowledge. In research methodologies, the choice between controlled and non-controlled studies significantly influences the study's reliability and accuracy.
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...

Accuracy and Errors in Hypothesis Testing

Accuracy and Errors in Hypothesis Testing

Hypothesis testing is a fundamental statistical tool that begins with the assumption that the null hypothesis H0 is true. During this process, two types of errors can occur: Type I and Type II. A Type I error refers to the incorrect rejection of a true null hypothesis, while a Type II error involves the failure to reject a false null hypothesis.
In hypothesis testing, the probability of making a Type I error, denoted as α, is commonly set at 0.05. This significance level indicates a 5%...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Comprehensive Analysis of <i>Artemisia argyi</i> H. Lév. & Vaniot Essential Oil: Chemical Composition, Antimicrobial Efficacy, Biofilm Disruption, and Insecticidal Potential.

Food science & nutrition·2026

Same author

Correction: Germ-line exon 21 EGFR V831H mutation in advanced NSCLC resistance to almonertinib: a case report.

Frontiers in oncology·2026

Same author

Ozone, Heat Shock, and Microwave Differentially Promote Nutritional Quality and Antioxidant Capacity of Sweet Corn.

Foods (Basel, Switzerland)·2026

Same author

Scap stabilizes PKM2 to promote glycolysis and enhance anti-fungal immunity in macrophages.

Cell reports·2026

Same author

Germ-line exon 21 EGFR V831H mutation in advanced NSCLC resistance to almonertinib: a case report.

Frontiers in oncology·2026

Same author

Sirtuin 3 Attenuates Acute Lung Injury by Decreasing Ferroptosis and Inflammation through Inhibiting Aerobic Glycolysis.

Biomedical and environmental sciences : BES·2025

Same journal

A Graph-based Framework for ABAC Policy Enforcement and Analysis.

Data and applications security and privacy XXXVIII : 38th Annual IFIP WG 11.3 Conference, DBSec 2024, San Jose, CA, USA, July 15-17, 2024, Proceedings. Annual IFIP WG 11.3 Working Conference on Data and Applications Security (38th : 202...·2025

See all related articles

Search research articles

Related Experiment Video

Updated: May 23, 2025

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index

Published on: January 8, 2020

Does Differential Privacy Prevent Backdoor Attacks in Practice?

Fereshteh Razmi¹, Jian Lou², Li Xiong¹

¹Emory University, Atlanta GA 30322, USA.

Data and Applications Security and Privacy XXXVIII : 38Th Annual IFIP WG 11.3 Conference, Dbsec 2024, San Jose, CA, USA, July 15-17, 2024, Proceedings. Annual IFIP WG 11.3 Working Conference on Data and Applications Security (38Th : 202

|May 22, 2025

Summary

This summary is machine-generated.

Differential Privacy techniques like DP-SGD, PATE, and Label-DP are evaluated for defending machine learning models against backdoor attacks. PATE shows effectiveness due to its bagging structure, while Label-DP requires careful tuning.

Keywords:

Backdoor Attack Differential Privacy Security

More Related Videos

Inducible and Reversible Dominant-negative DN Protein Inhibition

Inducible and Reversible Dominant-negative DN Protein Inhibition

Published on: January 7, 2019

The Use of Reverse Phase Protein Arrays RPPA to Explore Protein Expression Variation within Individual Renal Cell Cancers

The Use of Reverse Phase Protein Arrays RPPA to Explore Protein Expression Variation within Individual Renal Cell Cancers

Published on: January 22, 2013

Related Experiment Videos

Last Updated: May 23, 2025

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index

Published on: January 8, 2020

Inducible and Reversible Dominant-negative DN Protein Inhibition

Inducible and Reversible Dominant-negative DN Protein Inhibition

Published on: January 7, 2019

The Use of Reverse Phase Protein Arrays RPPA to Explore Protein Expression Variation within Individual Renal Cell Cancers

The Use of Reverse Phase Protein Arrays RPPA to Explore Protein Expression Variation within Individual Renal Cell Cancers

Published on: January 22, 2013

Area of Science:

Computer Science
Cybersecurity
Machine Learning

Background:

Differential Privacy (DP) is a privacy-preserving technique.
DP is increasingly used to defend machine learning (ML) models against poisoning and backdoor attacks.
DP-SGD has gained attention, but the efficacy of other DP methods against backdoor attacks needs thorough investigation.

Purpose of the Study:

To investigate the effectiveness of DP-SGD against backdoor attacks.
To examine PATE and Label-DP for the first time in the context of backdoor attacks.
To explore the role of DP algorithm components in defending against these attacks.

Main Methods:

Evaluating DP-SGD, PATE, and Label-DP against backdoor attacks on ML models.
Analyzing the impact of DP algorithm components on defense effectiveness.
Conducting experiments to assess the influence of hyperparameters and backdoor prevalence.

Main Results:

PATE demonstrates effectiveness against backdoor attacks, attributed to its teacher model bagging structure.
Hyperparameters and the number of backdoors significantly impact DP algorithm success.
Label-DP, despite weaker inherent privacy, can be effective with accurate tuning, rivaling other DP methods while preserving model accuracy.

Conclusions:

PATE offers a robust defense against backdoor attacks due to its inherent architecture.
Hyperparameter optimization is crucial for the efficacy of DP techniques, especially Label-DP.
DP methods, when properly tuned, can effectively defend ML models against backdoor attacks without compromising accuracy.