1. What is the minimum required set attributes of property attributes that can strongly imply the value of other associated attributes?

2. How can we predict associated attributes based on the minimum set of inputs through classification or recommendation techniques?

• Analyze the level of correlation between the various structural attributes of a building and identify key attributes that strongly imply the value of other attributes.

• Based on these correlations, build a model that is able to predict the structural attributes based on the minimum set of input.

Our data is collected from Partner’s property assessment reports, but we will be focusing mainly on Section 1, 4, and 5 of the reports.

• Section 1: Executive Summary 

  • 1. • Basic information about the property, such as build year etc..

• Section 4: Structure frame and Building envelope

• Section 5: Building plumbing, heating, electrical etc..

Using the data inputted in Section 1, we will be predicting variables in Section 4 and 5.

Two Layers:

1. Top down: Based on the expertise of our advisor from Partner, Bruce, we can infer what types of structural attributes are associated with other attributes.

2. More advanced goal: We can use preexisting collected data to identify patterns that aren’t already known to Partner to infer even more points.