The first task was about Simple regression. We have used several model Like Linear Regression, Bayessian Ridge, kernel Ridge and Random Forest Regressor. As the data was generated, each model did well. Among this, we choosed Linear Regression.
This task was on real life data. We had to predict output flow of a pump. We used shallow nural network. As the data is so much noisy(i.e. Faulty Sensor), a model which first try to filter out the wrong data, would do better. We used tensorflow.keras here.