Zillow Introduces New Networks Using the AWS Cloud

This technology is likely to advance the entire online real estate industry.

Salem-News.com

(SALEM, Ore.) - Zillow has long been an innovator and a leader in how the world sees and buys homes and a few years ago they even moved into the rental market, taking on a large-scale project to give future renters a better idea of what real life would in properties around the country.

Online Real Estate Technology

Zillow has been bringing together the world of real estate with cutting edge technology, and it’s helped them grow and succeed exponentially, even during times the real estate market itself wasn’t necessarily doing the same. Zillow even introduced Zestimates years ago, which used algorithms to determine the true value of properties featured on the site.

While the idea was great and loved by many consumers who visited the site, there were issues with accuracy on some properties, with owners complaining the Zestimate didn’t reflect the actual value of their home.

They haven’t given up on that sense of innovation, nor have they sidelined their efforts to bridge real estate and technology in new and exciting ways and are working on improving their Zestimates feature.

Neural Networks

News just broke that Zillow is now harnessing the power provided by the Amazon Web Services (AWS) Cloud to change the way people see the interior of homes and also how home values are estimated.

Zillow is calling the concept “neutral networks,” and essentially what it means is that the deep learning technology powered by AWS is replicating how the human brain sees and processes images around it.

It gets complicated, but essentially it’s bringing together collections of image pixels with what are being called “valuation signals.”

The network will let computers do something humans normally would, which is see images of a space and then look at individual features within that picture. Then, certain features of that home can be “tagged.” For example, if you were searching for a home and you walked in you might immediately notice the wide plank wood floors or the stone counter tops.

The neural networks created by Zillow can now do that as a human viewer would in many ways. A person can walk into a home and immediately process thousands of different points of data, and that’s why up until now, it’s been difficult to match the accuracy of computer-based estimates with the estimates that can be provided by an in-person appraisal.

Innovations Powered by AWS

With announcements like these, it’s not surprising that the cloud computing technology offered by AWS is so popular, and employers are on the constant search for employees with the AWS Certification.

Without these cloud computing services, it might not have been possible for Zillow to develop this technology, which is likely to advance the entire online real estate industry.

Increased Zestimate Accuracy

The result of these algorithms is meant to be a more accurate way to value properties. Zillow, with its announcement of its neural networks, said it has reduced its valuation error rate from 14% to 4.5%.

The company also said in its announcement that it hopes to continue developing the neutral networks to get even more accurate with valuations.

The goal from Zillow with the announcement of this news is that it will usher in a new era, where buyers will be able to search for properties in a way that’s as efficient and streamlined as possible, without any sacrifice in accuracy or relevance.

Zillow announced, with its news of the development of the neural network, that it should be in use by the first quarter of 2017.

Source: Salem-News.com Special Features Dept.