Synthetic Data for Deep Learning: Generate Synthetic Data for Decision Making and Applications with Python and R

ISBN-13: 9781484285862 / Angielski / Miękka / 2023 / 220 str.

Necmi Gürsakal; Sadullah Çelik; Esma Birişçi

Opis

Beginning-Intermediate user level

Data is the indispensable fuel that drives the decision making of everything from governments, to major corporations, to sports teams. Its value is almost beyond measure. But what if that data is either unavailable or problematic to access? That’s where synthetic data comes in. This book will show you how to generate synthetic data and use it to maximum effect.Synthetic Data for Deep Learning begins by tracing the need for and development of synthetic data before delving into the role it plays in machine learning and computer vision. You’ll gain insight into how synthetic data can be used to study the benefits of autonomous driving systems and to make accurate predictions about real-world data. You’ll work through practical examples of synthetic data generation using Python and R, placing its purpose and methods in a real-world context. Generative Adversarial Networks (GANs) are also covered in detail, explaining how they work and their potential applications.After completing this book, you’ll have the knowledge necessary to generate and use synthetic data to enhance your corporate, scientific, or governmental decision making.

What You Will Learn

• Create synthetic tabular data with R and Python
• Understand how synthetic data is important for artificial neural networks
• Master the benefits and challenges of synthetic data
• Understand concepts such as domain randomization and domain adaptation related to synthetic data generation

Who This Book Is For

Those who want to learn about synthetic data and its applications, especially professionals working in the field of machine learning and computer vision. This book will also be useful for graduate and doctoral students interested in this subject.

Spis treści

Chapter I: Introduction to Data 40 pages

Chapter Goal: The book section entitled "Data" aims to provide readers with information on the history, definition, and future of data storage, as well as the role that synthetic data can play in the field of computer vision. 

1.1. The History of Data

1.3. Definitions of Synthetic Data

1.4. The Lifecycle of Data

1.5. The Future of Data Storage

1.6. Synthetic Data and Metaverse

1.7. Computer Vision

1.8. Generating an Artificial Neural Network Using Package “nnet” in R

1.9. Understanding of Visual Scenes

1.10. Segmentation Problem

1.11. Accuracy Problems

1.12. Generative Pre-trained Transformer 3 (GPT-3) 

Chapter 2: Synthetic Data 40 pages

Chapter Goal: The purpose of this chapter is to provide information about synthetic data and how it can be used to benefit autonomous driving systems. Synthetic data is a term used to describe data that has been generated by a computer. 

2.1. Synthetic Data

2.2. A Brief History of Synthetic Data

2.3. Types of Synthetic Data

2.4. Benefits and Challenges of Synthetic Data

2.5. Generating Synthetic Data in A Simple Way

2.6. An Example of Biased Synthetic Data Generation

2.7. Domain Transfer

2.8. Domain Adaptation

2.9. Domain Randomization

2.10. Using Video Games to Create Synthetic Data

2.11. Synthetic Data And Autonomous Driving System

2.11.1. Perception

2.11.2. Localization

2.11.3. Prediction

2.11.4. Decision Making

2.12. Simulation in Autonomous Vehicle Companies

2.13. How to Make Automatic Data Labeling? 

2.14. Is Real-World Experience Unavoidable? 

2.15. Data for Learning Medical Images

2.16. Reinforcement Learning

2.17. Self-Supervised Learning

Chapter 3: Synthetic Data Generation with R..... 55 pages

Chapter Goal: The purpose of this book section is to provide information about the content and purpose of synthetic data generation with R. Synthetic data is generated data that is used to mimic real data. There are many reasons why one might want to generate synthetic data. For example, synthetic data can be used to test data-driven models when real data is not available. Synthetic data can also be used to protect the privacy of individuals in data sets.

3.1. Basic Functions Used In Generating Synthetic Data

3.1.1. Creating a Value Vector from a Known Univariate Distribution

3.1.2. Vector Generation from a Multi-levels Categorical Variable

3.1.3. Multivariate

3.1.4. Multivariate (with correlation) 

3.2. Multivariate Imputation Via Mice Package in R

3.2.1. Example of MICE

3.3. Augmented Data

3.4. Image Augmentation Using Torch Package

3.5. Generating Synthetic Data with The "conjurer" Package in R

3.5.1. Create a Customer

3.5.2. Create a Product

3.5.3. Creating Transactions

3.5.4. Generating Synthetic Data

3.6. Generating Synthetic Data With “Synthpop” Package In R

3.7.    Copula

3.7.1. t Copula

3.7.2. Normal Copula

3.7.3. Gaussian Copula

Chapter 4: GANs.... 15 pages

Chapter Goal: This book chapter aims to provide information on the content and purpose of GANs. GANs are a type of artificial intelligence that is used to generate new data that is similar to the training data. This is done by training a generator network to produce data that is similar to the training data. The generator network is trained by using a discriminator network, which is used to distinguish between the generated data and the training data. 

4.1.      GANs

4.2.    CTGAN

4.3. SurfelGAN

4.4. Cycle GANs

4.5. SinGAN

4.6. DCGAN

4.7. medGAN

4.8. WGAN

4.9. seqGAN

4.10. Conditional GAN

Chapter 5: Synthetic Data Generation with Python.... 40 pages

Chapter Goal: The purpose of this chapter is to provide information about the methods of synthetic data generation with Python. Python is a widely used high-level programming language that is known for its ease of use and readability. It has a large standard library that covers a wide range of programming tasks.

5.1. Data Generation with Know Distribution

5.2. Synthetic Data Generation in Regression Problem

5.3. Gaussian Noise Apply to Regression Model

5.4. Friedman Functions and Symbolic Regression

5.5. Synthetic data generation for Classification and Clustering Problems

5.6. Clustering Problems

5.7. Generation Tabular Synthetic Data by Applying GANs

O autorze

Necmi Gürsakal is a statistics professor at Mudanya University, where he transfers his experience and knowledge to his students. Before that, he worked as a faculty member at the Bursa Uludag University Econometrics Department for more than 40 years. Necmi has many published Turkish books and English and Turkish articles on data science, machine learning, artificial intelligence, social network analysis, and big data. In addition, he has served as a consultant to various business organizations.

Sadullah Çelik completed his undergraduate and graduate education in mathematics and his doctorate in statistics. He has written numerous Turkish and English articles on big data, data science, machine Learning, Generative Adversarial Networks (GANs), multivariate statistics, and network science. He has authored three books: Big Data, R Applied Linear Algebra for Machine Learning and Deep Learning, and Big Data and Marketing. Sadullah is currently working as Research Assistant at Aydın Adnan Menderes University, Nazilli Department of Economics and Administrative Sciences, and Department of International Trade and Finance.

Esma Birişçi is a programmer, statistician, and operations researcher with more than 15 years of experience in computer program development and five years in teaching students. She developed her programming ability while studying for her bachelor degree, and knowledge of machine learning during her master degree program. She completed her thesis about data augmentation and supervised learning. Esma transferred to Industrial Engineering and completed her doctorate program on dynamic and stochastic nonlinear programming. She studied large-scale optimization and life cycle assessment, and developed a large-scale food supply chain system application using Python. She is currently working at Bursa Uludag University, Turkey, where she transfers her knowledge to students. In this book, she is proud to be able to explain Python's powerful structure.

Blurb

Data is the indispensable fuel that drives the decision making of everything from governments, to major corporations, to sports teams. Its value is almost beyond measure. But what if that data is either unavailable or problematic to access? That’s where synthetic data comes in. This book will show you how to generate synthetic data and use it to maximum effect.

Synthetic Data for Deep Learning begins by tracing the need for and development of synthetic data before delving into the role it plays in machine learning and computer vision. You’ll gain insight into how synthetic data can be used to study the benefits of autonomous driving systems and to make accurate predictions about real-world data. You’ll work through practical examples of synthetic data generation using Python and R, placing its purpose and methods in a real-world context. Generative Adversarial Networks (GANs) are also covered in detail, explaining how they work and their potential applications.

After completing this book, you’ll have the knowledge necessary to generate and use synthetic data to enhance your corporate, scientific, or governmental decision making.

What You Will Learn

• Create synthetic tabular data with R and Python
• Understand how synthetic data is important for artificial neural networks
• Master the benefits and challenges of synthetic data
• Understand concepts such as domain randomization and domain adaptation related to synthetic data generation