Deep Learning Approaches with Regularization, Generative Transformation, and Residual Learning in Biomedical and Material Sciences = 딥러닝 기반 정규화, 생성 변환, 잔차 학습을 활용한 바이오 및 재료 과학 분야에서의 연구

다국어 초록 (Multilingual Abstract)

This thesis covers a variety of deep learning approaches with regularization, style-transfer generation, and residuals applied to biomedical and material sciences. The first study introduces a novel deep neural network architecture, deep learning-based cancer metastasis estimator (DeepCME), designed to predict breast cancer metastasis using gene expression. However, the problem of overfitting occurs frequently while training deep learning models using gene expression data because they contain a large number of genes, and the sample size is rather small. To address this problem, several regularization methods are implemented, such as L1 penalty, batch normalization, and dropout. DeepCME demonstrates superior performance, achieving the highest average area under curve scores compared with five baseline models. Additionally, significant genes associated with breast cancer metastasis are identified, highlighting the potential for cost-effective and time-efficient clinical applications. The second study introduces the combined transformation GAN (ctGAN), a style-transfer generative model designed to address existing challenges in the implementation of deep learning models for analyzing gene expression. Previous models have limited applicability for clinical purposes, but ctGAN enables the combined transformation of gene expression and survival data and improves survival analysis by augmenting data through style transformations. ctGAN demonstrates high plausibility in data generation based on distribution and clustering evaluations. The proposed method may enable predictions regarding the likelihood of a patient with breast cancer developing other types of cancer and responding differently to various treatment methods. The third study focuses on estimating Single-Parabolic Band (SPB) parameters in thermoelectric materials using residual-based ensemble model. Estimating SPB parameters is significant for understanding the band structure and predicting the theoretical performance of thermoelectric materials. However, the derivation of these band parameters involves complex calculations, including Fermi integrals, posing substantial challenges due to their computational complexity. To address this problem, this study proposes DeepSPB which is an ensemble model with deep neural network and machine learning model. Deep neural networks are trained to predict band parameters, including density of states effective mass, nondegenerate mobility, and reduced Fermi level. After that, an additional machine learning model leveraging residuals was implemented to adjust predictions by estimating error rates, effectively reducing the prediction error. The performance of the models was validated using pseudo data and experimental data for ZnSb decorated with Ag and Cu.
The three methods described in this paper demonstrate considerable potential for application in various industries. The first two studies could improve metastatic cancer prediction and support cancer treatment and prevention in the medical field, benefiting public healthcare. The final study could facilitate material optimization in the semiconductor industry by analyzing band parameter variations with doping concentrations, reducing time and costs.

목차 (Table of Contents)

• TABLE OF CONTENTS
• ABSTRACT i
• 국문 초록 iv
• ACKNOWLEDGEMENTS viii
• TABLE OF CONTENTS x
• LIST OF TABLES xiii
• LIST OF FIGURES xiv
• CHAPTER 1. INTRODUCTION 1
• 1.1 Interdisciplinary Deep Learning Approaches 1
• 1.2 Organization of the thesis 4
• CHAPTER 2. Deep Learning Model with L1 Penalty for Predicting Breast Cancer Metastasis Using Gene Expression Data 5
• 2.1 Introduction 7
• 2.2 Methods 10
• 2.2.1 Model architecture 11
• 2.2.1.1 MLPs 12
• 2.2.1.2 BN 14
• 2.2.1.3 Dropout 16
• 2.2.1.4 L1 penalty 17
• 2.2.2 Gene score 19
• 2.2.3 Experimental settings 20
• 2.3 Results 22
• 2.4 Conclusions 30
• CHAPTER 3. ctGAN: Combined Transformation of Gene Expression and Survival Data with Generative Adversarial Network 32
• 3.1 Introduction 34
• 3.2 Material and methods 37
• 3.2.1 Model architecture 37
• 3.2.2 Model training and parameter setting 41
• 3.2.3 C-index 43
• 3.2.4 Survival analysis method 44
• 3.2.5 Dataset 45
• 3.2.6 Gene selection 46
• 3.2.7 Overview of style transfer workflow 50
• 3.2.8 Architecture and training of other frameworks 52
• 3.3 Results 54
• 3.3.1 Evaluation of survival analysis enhancement 54
• 3.3.2 Style transfer validation 60
• 3.3.3 ctGAN with SCAN-B dataset 67
• 3.4 Conclusion 69
• CHAPTER 4. Estimating Single-Parabolic Band Parameters in Thermoelectric Materials using an Ensemble model with Residual Learning 71
• 4.1 Introduction 73
• 4.2 Experimental details 77
• 4.2.1 Synthesis of ZnSb and decoration of Au, Cu, and excess 77
• 4.2.2 Thermoelectric properties characterization 78
• 4.2.3 Single Parabolic Band (SPB) modeling 78
• 4.2.4 Model architecture 80
• 4.2.5 Ensemble model and residual learning 81
• 4.2.6 Dataset 82
• 4.2.7 Experimental settings 83
• 4.3 Results 84
• 4.3.1 Pseudo data 84
• 4.3.2 ZnSb decoration with Cu and Ag 85
• 4.4 Conclusions 88
• CHAPTER 5. CONCLUSION 90
• REFERENCES 91