Assignment Task: Tony is a realtor at Century 21 in the Chicago, Illinois metropolitan area. She manages a large real estate portfolio of different housing properties in the region. She has assembled a database of information both on properties she and her real estate firm manage or have previously bought and sold. Tony has hired you to analyze her data and provide insights into how the various features of the properties in her dataset affect the selling price of the housing properties that she manages. She would like you to formulate, train, and consider predictive models for the selling price, using several different machine-learning techniques(regression/regularized regression), a single decision tree, a random forest, a gradient boosting machine, a neural network, and a support vector machine. She would like you to present a technical report explaining your findings. It is up to you to determine what aspects of the analysis you would like to focus on, but some possibilities include looking at the overall predictive error, differences between training and testing errors, measures of feature importance, and the advantages and disadvantages of different modeling techniques with regards to this problem. Be sure to make a judgment on which of your models is the optimal one for this problem, and why. Your report should include an executive summary that is comprehensible to a non-specialist such as Tony.
This analysis must be done in R (the programming language)
The data used is ames_housing data.
The algorithm can be any three of the above-mentioned algorithm.