SyntheVAEiser: augmenting traditional machine learning methods with VAE-based gene expression sample generation for improved cancer subtype predictions
Brian Karlberg, Raphael Kirchgaessner, Jordan Lee et al. · Genome biology · 2024
The accuracy of machine learning methods is often limited by the amount of training data that is available. We proposed to improve machine learning training regimes by augmenting datasets with synthetically generated samples. We present a method for synthesizing gene expression samples and test the system's capabilities for improving the accuracy of categorical prediction of cancer subtypes. We developed SyntheVAEiser, a variational autoencoder based tool that was trained and tested on over 8000 cancer samples. We have shown that this technique can be used to augment machine learning tasks and increase performance of recognition of underrepresented cohorts.
Cited by 6 works. doi:10.1186/s13059-024-03431-3 Open-access PDF
Authors
References
- SMOTE: Synthetic Minority Over-sampling Technique
- Integrated genomic analyses of ovarian carcinoma
- A survey on Image Data Augmentation for Deep Learning
- Supervised Risk Predictor of Breast Cancer Based on Intrinsic Subtypes
- An empirical comparison of supervised learning algorithms
- Comprehensive genomic characterization of head and neck squamous cell carcinomas
- Machine learning applications in genetics and genomics
- Comprehensive Molecular Portraits of Invasive Lobular Breast Cancer
- A guide to machine learning for biologists
- SMOTE for high-dimensional class-imbalanced data