This unique book aims to address the challenges of machine learning statistics, emphasize the importance of performance valuation by appropriate metrics and independent data, and describe several important concepts of machine learning statistics, such as feature engineering and overfitting.