Wednesday, April 26, 2017

Is Machinery & Equipment Appraising Compatible With Big Data?

1.
Big Data.  Its been a buzz word in the Technology community for several years now, and it is even being joked about in the general population.  I recently attended a webinar hosted by Iron Planet where the topic was Predicting the Future of Equipment Machine Pricing.  Naturally I was interested as a Machinery & Equipment Appraiser.

The webinar was basically about 3 features Iron Planet is working on: virtual storefronts, market place direct, and deep learning pricing data.  While everyone who attended may be interested for multiple reasons, my attention was on the deep learning pricing data.  This is code for Iron Planet putting to use the glut of data they have been collecting over the years, or putting to use their own Big Data.

Several years ago I wrote an article on the residential mortgage industry's rapid adoption of Big Data know as Collateral Underwriter.  Keep in mind this article was for a different industry and was spear headed by the lending/government portion of the transaction.  In hind sight, the concerns I had back then were well founded as we've seen testimonies from Capital Hill that confirm our distrust of the output.  Just ask lenders in Canada how there version of the Collateral Underwriter is doing.  The problem with CU, and any predictive model, is the output is only as good as the data going in.  The CU is not peer reviewed, and the quality of its data has been in question since day one.  Recently, an interview on Voice of Appraisal revealed how lenders have figured out to manipulate the portal's risk scores, crazy!

However, with Iron Planet's model, I think they are going about it in a healthy direction.  First, they have put together a crack team that draws from successful companies like: Paypal and Ebay.  Both of these companies have mastered predictive analytics, finances, and buy/sell in a big way.  Second, the data collection has been taking place over a longer period of time allowing for data to come from both ends of the Great Recession.  Lastly, this is an industry driven initiative, not a government backed one, which I'm not a big fan of.

In closing, I'm looking forward to what Iron Planet does with their deep learning pricing data.  I'm also looking forward to how it could benefit the Machinery & Equipment Appraisal field.




1. http://www.learntek.org/category/big-data-and-hadoop/