Research helps home buyers better predict property prices
The ability to predict house prices more accurately has become a reality, thanks to RMIT-developed software.
Helping home buyers more accurately predict the selling price of properties
Vic Ciesielski
Prospective home buyers have often found it difficult to accurately predict the selling price of a property and this can add unwanted pressures as they consider such a major investment.
RMIT researchers developed a software program, REALas that predicts within 5 per cent of the final sales price and is now the most accurate property sales price predictor on the market.
The foundations for the program stem from an algorithm created by RMIT data mining expert and lead researcher, Vic Ciesielski.
Using data mining techniques and predictive algorithms, Ciesielski was working on another property application when he found that he could predict the property sales prices with surprising accuracy.
The program looks at property information from a range of sources to achieve more accurate results.
To make the products marketable, a team of students were employed to design a web interface for the software. RMIT alumnus Josh Rowe was then selected to take the project further.
With Rowe as CEO, the REALas program went on to win the Westpac Innovation Challenge in 2014 before the business was sold to ANZ in 2017.
The REALas program continues to grow with 12 people employed to run the system and manage the data from the ANZ head office in Collins Street.