Hemorrhagic shock (HS) potentially impacts the chance of survival in most traumatic injuries. Thus, it is highly desirable to maximize the survival rate in cases of blood loss by predicting the occurrence of hemorrhagic shock with biomedical signals. Since analyzing one physiological signal may not enough to accurately predict blood loss severity, two types of physiological signals - Electrocardiography (ECG) and Transcranial Doppler (TCD) - are used to discover the degree of severity. In this study, these degrees are classified as mild, moderate and severe, and also severe and non-severe. The data for this study were generated using the human simulated model of hemorrhage, which is called lower body negative pressure (LBNP). The analysis is done by applying discrete wavelet transformation (DWT). The wavelet-based features are defined using the detail and approximate coefficients and machine learning algorithms are used for classification. The objective of this study is to evaluate the improvement when analyzing ECG and TCD physiological signals together to classify the severity of blood loss. The results of this study show a prediction accuracy of 85.9% achieved by support vector machine in identifying severe/non-severe states.