Real Estate


Jack Ma warns about dangers of artificial intelligence

#artificialintelligence

"Artificial intelligence may take a lot of jobs away. If we don't give simple and easy technology products for small businesses, most of them can't survive in 10 years. If small businesses can't survive, we can't survive," said Ma, speaking in Detroit at the Gateway 17 conference for entrepreneurs. According to Goldman Sachs, professions like truckers, secretaries, cashiers, bank tellers, waiters and real estate agents could be replaced by artificial intelligence in the near-term.


Stacking models for improved predictions: A case study for housing prices

@machinelearnbot

I will use three different regression methods to create predictions (XGBoost, Neural Networks, and Support Vector Regression) and stack them up to produce a final prediction. I trained three level 1 models: XGBoost, neural network, support vector regression. Graphically, once can see that the circled data point is a prediction which is worse in XGBoost (which is the best model when trained on all the training data), but neural network and support vector regression does better for that specific point. For example, below are the RMSE values on the holdout data (rmse1: XGBoost, rmse2: Neural Network, rmse3: Support Vector Regression), for 20 different random 10-folds created.


What is machine learning debt?

#artificialintelligence

For a practical guide to integrate and test machine learning algorithms, check out Matthew Kirk's Thoughtful Machine Learning with Python. In 2007 when the housing market crashed, a lot of the stock market's correlations changed substantially. For instance, a machine learning project that has issues with data and math debt will compound in complexity, and together, both types of debt will make it a difficult project to maintain. For a practical guide to integrate and test machine learning algorithms, check out Matthew Kirk's Thoughtful Machine Learning with Python.


How blockchain can improve the mortgage process

#artificialintelligence

The pressure is particularly strong with FinTechs like US online lender Rocket Mortgage and UK digital mortgage broker Trussle creating a completely digital experience for prospective home buyers. This was the exact focus of Synechron's blockchain accelerator for mortgage financing and processing where we re-architected business processes and developed an accelerator application to help banks leap-frog the innovation stages needed to embrace this type of technology. Additionally, a public blockchain for real estate title, deeds, planning permissions, mortgage registry and other public records associated with the real estate assets could provide a second powerful application to further enhance these processes. In fact, in Dubai, Synechron recently won a leading position in a hackathon for its Land Registry team's blockchain submission which built an application to automatically generate title deeds on the blockchain.


Machine learning that could make realtors extinct (VB Live)

#artificialintelligence

At REX, a real estate service platform, the complex problem seems deceptively simple at first glance. They get the info from traditional consumer marketing companies, which means they're capturing an extraordinary variety of information. That data gets poured into what Barkett calls classic supervised machine learning, which means the algorithm is actually tuned to reach a specific outcome, like signing up to list their house or actually picking up when someone from REX calls. "So rather than try to speculate about what's predictive, and rather than try to do a simple, one-time static analysis of variables to figure out which is interesting, we just continuously run a machine learning process and let the machine decide which variables are predictive of those outcomes," Barkett says.


How to Solve the New $1 Million Kaggle Problem - Home Value Estimates

@machinelearnbot

More specifically, I provide here high-level advice, rather than about selecting specific statistical models or algorithms, though I also discuss algorithm selection in the last section. If this is the case, an easy improvement consists of increasing value differences between adjacent homes, by boosting the importance of lot area and square footage in locations that have very homogeneous Zillow value estimates. Then for each individual home, compute an estimate based on the bin average, and other metrics such as recent sales price for neighboring homes, trend indicator for the bin in question (using time series analysis), and home features such as school rating, square footage, number of bedrooms, 2- or 3-car garage, lot area, view or not, fireplace(s), and when the home was built. With just a few (properly binned) features, a simple predictive algorithm such as HDT (Hidden Decision Trees - a combination of multiple decision trees and special regression) can work well, for homes in zipcodes (or buckets of zipcodes) with 200 homes with recent historical sales price.


Senior Machine Learning Engineer posted by HomeAway.com on DigitalMediaJobsNetwork.com

#artificialintelligence

Want to be part of a team where that all comes together? Want a chance to drive an extraordinary experience for travelers by connecting them with millions of vacation rental properties? Passionate about building and putting complex models into production to solve relevance and personalization problems, great at model implementation, testing and evaluation? If so, then HomeAway has a Machine Learning Engineer role for you and would like to talk to you!


How to Solve the New $1 Million Kaggle Problem - Home Value Estimates

#artificialintelligence

More specifically, I provide here high-level advice, rather than about selecting specific statistical models or algorithms, though I also discuss algorithm selection in the last section. If this is the case, an easy improvement consists of increasing value differences between adjacent homes, by boosting the importance of lot area and square footage in locations that have very homogeneous Zillow value estimates. Then for each individual home, compute an estimate based on the bin average, and other metrics such as recent sales price for neighboring homes, trend indicator for the bin in question (using time series analysis), and home features such as school rating, square footage, lot area, view or not, and when the home was built. With just a few (properly binned) features, a simple predictive algorithm such as HDT (Hidden Decision Trees - a combination of multiple decision trees and special regression) can work well, for homes in zipcodes (or buckets of zipcodes) with 200 homes with recent historical sales price.


Predicting House Sales

@machinelearnbot

To impute the value of a house for years in which it did not change hands, I applied the log returns of the Case Shiller Charlotte Home Price Index to the most recent sale price. Therefore, if random forests greatly outperformed logistic regression, I could prioritize feature engineering to help improve logistic regression's predictiveness while still retaining its descriptiveness. By calibrating the cutoff on the logit function to yield a false positive rate of 0.5, we can greatly increase the true positive rate for a subset of households, as depicted in the below tables: If a real estate broker were to use this model, they would be 60% more likely to distinguish prospects from non-prospects. Historical patterns could instead be used to look for inflection points where next year's sales may be less predictable based on this year's data due to a potential change in macro factors.


Alphabet's Sidewalk Labs Eyes Toronto for Its Digital City

@machinelearnbot

Sidewalk Labs LLC, the urban innovation unit of Page's Alphabet Inc., has applied to develop a 12-acre strip in downtown Toronto, responding to a recent city agency request for proposals, according to two people familiar with the plans. Before applying in Toronto, Sidewalk Labs discussed creating a district in Denver and Detroit with Alphabet executives, according to the people. In a speech last week at the Smart Cities NYC conference, Sidewalk Labs Chief Executive Officer Dan Doctoroff said the firm is exploring development of a "large-scale district." So far, the most visible project is LinkNYC, a network of ad-supported Wi-Fi kiosks in New York City run by Intersection, a Sidewalk Labs investment.