Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Variability: Analysis01:11

Variability: Analysis

288
Measures of variability are statistical metrics that reveal the dispersion pattern within a dataset. They are pivotal in biostatistics, providing insights into the heterogeneity within health and biological data. Variability signifies the degree to which data points diverge from one another, helping researchers understand the potential range of values and associated uncertainty within the data.
The range is a simple measure of variability, indicating the difference between the highest and...
288
Random Variables01:09

Random Variables

16.7K
A random variable is a single numerical value that indicates the outcome of a procedure. The concept of random variables is fundamental to the probability theory and was introduced by a Russian mathematician, Pafnuty Chebyshev, in the mid-nineteenth century.
Uppercase letters such as X or Y denote a random variable. Lowercase letters like x or y denote the value of a random variable. If X is a random variable, then X is written in words, and x is given as a number.
For example, let X = the...
16.7K
Variance01:15

Variance

11.4K
The deviations show how spread out the data are about the mean. A positive deviation occurs when the data value exceeds the mean, whereas a negative deviation occurs when the data value is less than the mean. If the deviations are added, the sum is always zero. So one cannot simply add the deviations to get the data spread. By squaring the deviations, the numbers are made positive; thus, their sum will also be positive.
The standard deviation measures the spread in the same units as the data....
11.4K
Randomized Experiments01:13

Randomized Experiments

8.6K
The randomization process involves assigning study participants randomly to experimental or control groups based on their probability of being equally assigned. Randomization is meant to eliminate selection bias and balance known and unknown confounding factors so that the control group is similar to the treatment group as much as possible. A computer program and a random number generator can be used to assign participants to groups in a way that minimizes bias.
Simple randomization
Simple...
8.6K
Survival Tree01:19

Survival Tree

247
Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
 Building a Survival Tree
Constructing a...
247
Data Collection by Experiments01:13

Data Collection by Experiments

26.6K
Data collection is a systematic method of obtaining, observing, measuring, and analyzing accurate information. An experimental study is a standard method of data collection that involves the manipulation of the samples by applying some form of treatment prior to data collection. It refers to manipulating one variable to determine its changes on another variable. The sample subjected to treatment is known as “experimental units.”
An example of the experimental method is a public...
26.6K

You might also read

Related Articles

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

Sort by
Same author

Association of a standardized nursing protocol with postoperative complications following open radical gastrectomy: a retrospective cohort study.

Open medicine (Warsaw, Poland)·2026
Same author

Comparative safety evaluation of ADAS-Equipped electric and gasoline vehicles using real-world crash data.

Accident; analysis and prevention·2026
Same author

When does visual distraction become dangerous in car-following? Evidence from naturalistic driving study data with causal inference on time-to-collision and braking intensity.

Accident; analysis and prevention·2026
Same author

Determinants influencing risks in e-bike cyclists under mix traffic condition: a partially constrained random parameters approach using experimental study data.

Accident; analysis and prevention·2025
Same author

Segment level safety analysis using lane-changing behavior and driving volatility features from connected vehicle trajectories.

Scientific reports·2025
Same author

Assessing the safety effectiveness of advanced driver assistance systems.

Journal of safety research·2025
Same journal

The grand rapids dip could it be real?

Accident; analysis and prevention·2026
Same journal

A temporal causal deep learning for real-time traffic risk prediction on freeway merging areas.

Accident; analysis and prevention·2026
Same journal

Modeling road-segment-level speeding risk of new energy vehicle taxis using a multistage framework with spatial spillover, endogeneity, and nonlinear effects.

Accident; analysis and prevention·2026
Same journal

Role of streetscape feature in pedestrian safety: A modified multi-level multiple membership model.

Accident; analysis and prevention·2026
Same journal

Assessing autonomous driving performance and environmental influencing factors using real-world operational trajectory data.

Accident; analysis and prevention·2026
Same journal

Multi-scale modeling of electric vehicle fatal crash risk: uncovering spatial heterogeneity and infrastructure-land use coupling mechanisms.

Accident; analysis and prevention·2026
See all related articles

Related Experiment Video

Updated: Nov 24, 2025

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.5K

Crash data augmentation using variational autoencoder.

Zubayer Islam1, Mohamed Abdel-Aty1, Qing Cai1

  • 1Department of Civil, Environmental and Construction Engineering, University of Central Florida, Orlando, FL, 32816, USA.

Accident; Analysis and Prevention
|December 28, 2020
PubMed
Summary
This summary is machine-generated.

This study introduces a variational autoencoder (VAE) for data augmentation to address imbalanced crash datasets. VAE successfully generates statistically similar crash data, outperforming other methods like SMOTE and ADASYN in prediction models.

Keywords:
Crash predictionData augmentationVariational autoencoder

More Related Videos

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

853

Related Experiment Videos

Last Updated: Nov 24, 2025

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.5K
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

853

Area of Science:

  • Machine Learning
  • Data Science
  • Artificial Intelligence

Background:

  • Real-world datasets, such as crash data, are often extremely imbalanced, with significantly fewer crash events than non-crash events.
  • This imbalance poses a challenge for machine learning algorithms, leading to poor performance in identifying minority class events.
  • The dataset used in this study contained only 625 crash events against over 6.5 million non-crash events.

Purpose of the Study:

  • To develop and evaluate a novel data augmentation technique using a variational autoencoder (VAE) to address class imbalance in crash datasets.
  • To generate synthetic crash data that is statistically similar to real-world crash data.
  • To compare the effectiveness of VAE-generated data against other oversampling techniques and generative frameworks for improving crash prediction models.

Main Methods:

  • A variational autoencoder (VAE) was employed to encode crash and non-crash events into a latent space.
  • The trained VAE model was used to sample from the latent space representing crash data, generating synthetic crash events.
  • Statistical tests (t-test, Levene-test, Kolmogrove Smirnov) were conducted to compare the generated data with real crash data.
  • Performance was evaluated using crash prediction models (Logistic Regression, Support Vector Machine, Artificial Neural Network) trained on data augmented by VAE, SMOTE, ADASYN, and GANs.

Main Results:

  • The VAE model successfully separated crash and non-crash events in the latent space.
  • Statistical analyses confirmed that the VAE-generated data is statistically similar to the real crash data.
  • Crash prediction models utilizing VAE-augmented data demonstrated superior performance compared to models using SMOTE and ADASYN.
  • Specificity improved by 8% (VAE-LR vs. SMOTE) and 4% (VAE-SVM vs. SMOTE); sensitivity improved by 6% (VAE-LR vs. ADASYN) and 5% (VAE-SVM vs. ADASYN).
  • VAE-generated data helped mitigate the overfitting issues commonly observed with SMOTE and ADASYN.

Conclusions:

  • Variational autoencoders offer a powerful and effective data augmentation strategy for highly imbalanced datasets like crash data.
  • VAE-generated data significantly enhances the performance of machine learning models in crash prediction tasks.
  • The VAE approach provides a flexible alternative to existing oversampling techniques, improving model accuracy and reducing overfitting.