This seminar is an opportunity to get an experience in data-driven mathematical models which have been popular in both industry and academic research over the past two decades. Indeed, in many real-world applications, the physical/chemical/biological process of interest is so complex that its mathematical model using solely physical/chemical/biological laws is just partially available (because of limited scientific results in the relevant area). Even when the model is available it is challenging to determine modeling parameters. Also, dealing with the model computationally is so expensive, especially in real-time applications. Fortunately, in many applications, one could hope to assimilate measurement data as a leverage for constructing mathematical models which are computationally efficient to understand and explore the process. Data-driven mathematical models employ methods ranging from purely data-based approaches coming from statistical and machine learning techniques to those which also encode underlying physical/chemical/biological laws to construct mathematical models governing a given data set.