A Neural Network Based Method with Transfer Learning for Genetic Data Analysis
This work addresses genetic data analysis, but is incremental as it applies an existing technique (transfer learning) to a new domain.
The authors applied transfer learning to expectile neural networks for genetic data analysis, demonstrating performance improvements on two real datasets compared to using the neural networks without transfer learning.
Transfer learning has emerged as a powerful technique in many application problems, such as computer vision and natural language processing. However, this technique is largely ignored in application to genetic data analysis. In this paper, we combine transfer learning technique with a neural network based method(expectile neural networks). With transfer learning, instead of starting the learning process from scratch, we start from one task that have been learned when solving a different task. We leverage previous learnings and avoid starting from scratch to improve the model performance by passing information gained in different but related task. To demonstrate the performance, we run two real data sets. By using transfer learning algorithm, the performance of expectile neural networks is improved compared to expectile neural network without using transfer learning technique.