In the age of growing presence of social media and online news sources, it is becoming difficult to distinguish between the real and the fake news. Thus the credibility detection of a news is of high importance such as social or political importance. However, assessing the credibility of the news is still challenging especially when the news is not entirely fake. Therefore, this thesis aims to assess the credibility of the news based on the continuous scale. The proposed approach is to solve this problem by combining existing machine learning models for fake news detection and create a hybrid model to utilize various advantages of machine learning models. The datasets used for this work consists of the news articles related to the health and engagements on Twitter. The model is expected to automatically assign the credibility score of news from health and engagements.