Visualization of Feature Engineering Strategies for Predictive Analytics

1. Introduction

As explained in Buytendijk and Trepanier’s work, (2010) predictive models can find relationship between outcome and dependent variables. There are six phases for predictive analytics process. In the initial phase, project is defined with outcomes, objectives, scope and the deliverables from the project. In the next phase, data is collected from various sources and is analyzed. This analysis requires strategies for preprocessing such as data cleaning, transformation and data modeling so that useful data is extracted for further processing. Subsequently validate the initial hypothesis using statistical models. The next phase is predictive modeling for forecasting the future. Results after implementation can be deployed for using it in the day to day decision making. The last phase is, monitoring the model in order to ensure that it is providing the expected results. Performance of computing layer of our framework is described in Babu, Vasavi, and Nagarjuna (2017) (Babu, & Vasavi, 2018). This layer finds, which of the algorithms such as Artificial Neural Networks (ANN), Auto Regression algorithm (ARX) and Gaussian process (GP) is better for income tax dataset to identify fraud in the projected tax values. Feature Engineering is the way toward changing crude information into features that better represents to the basic issue to the predictive models, bringing about enhanced model accuracy on unseen information. The execution of machine learning strategies is intensely dependent on the selection of data representation on which they are connected. Hence, a great part of the actual effort in deploying machine learning. Calculations goes into the plan of preprocessing pipelines and information changes that outcome in are introduction of the information that can support successful machine learning. Feature selection is the process of finding suitable features to be used in the model building. Feature selection techniques are used for four phases:

• 1.

  Simplification of models so as to interpret by scientists/clients;

• 2.

  Shorter training time;

• 3.

  Execration of dimensionality;

• 4.

  Enhanced generalization by decreasing over fitting.