In this chapter, we took our predictive analytics skillset to a notch higher. You learned and practically implemented sophisticated cutting-edge machine learning and deep learning algorithms to improve our results in the predictive power. We studied ensemble modeling techniques in machine learning such as random forest and extreme gradient boosting, xgboost. You also learned the basics of neural networks and deep learning using Multilayered Perceptrons, that is, MLP. In the overall exercise, we achieved better and improved results for our use case to predict the end quality of the detergent before the manufacturing process. We built a valuable solution for John and his team with an opportunity where they could take immediate actions to mitigate the bad quality produce and reduce the overall losses by 16%.
In the next chapter, we will reinforce our problem solving and decision science skills by solving another IoT use case in a fast-track mode. We'll revisit the decision science journey...