In this chapter, we discussed various methods of performing dimensionality reduction in a sample dataset. Removing redundant features not only improves model accuracy, but also saves computational effort and time. From business user's point of with less number of dimensions, it is more intuitive to build strategies than to focus on large number of features. We discussed which technique to use where and what the data requirement for each of the methods is. The reduction in dimensions also provides meaningful insights into large datasets. In the next chapter, we are going to learn about neural network methods for classification, regression, and time series forecasting.