Heat transfer processes play a major role for the developments in many areas of science and engineering and performance of many energy systems. Recent advances in experimental and computational methods are increasing the quantity and complexity of generated data in heat transfer problems. This study presents an application of Artificial Intelligence (AI) in heat transfer area by using Deep Learning (DL), which is one of the Machine Learning (ML) methods and is based on the Artificial Neural Network (ANN). The present work considers both applications in forward modelling DL and inverse modelling DL approaches for heat transfer natural convection problems. For the inverse problem, the forward problems are solved first to create a database. This database is then used to train the machine learning algorithms. The trained algorithm is then used to determine the boundary conditions of a problem from assumed measurements. The correctness of the algorithms is checked by comparing the statistical evaluation metrics, such as accuracy and loss on in/out of sample data. Interpretability tools such as SHAP (SHapley Additive exPlanations) helps to better understand how machine learning models work especially when moving towards more complex systems that process large amount of data. Finally, the effectiveness of using ML algorithms for predicting the thermal performance of heat exchangers through Colburn and Fanning friction factors is presented.