Subnet 48
Nextplace AI
Nextplace AI is a real estate subnet using AI to estimate home prices and track market trends.
SN48 : NextPlace AI
| Subnet | Description | Category | Company |
|---|---|---|---|
| SN48 : NextPlace AI | Property price prediction | Predictive systems | Nickel5 |
Welcome to Nextplace AI the first real estate subnet within the Bittensor Ecosystem. Harnessing the power of AI through the Bittensor framework, Nextplace is developing the ultimate real estate market negotiator. With their platform, you can discover the estimated selling price of your home, identify housing trends in real estate markets across the United States, and earn TAO by deploying your own AI models in their subnet, among many other features.
They gather data from the most dynamic housing markets in the USA, including cities like New York, San Francisco, and Seattle. their subnet miners utilize AI models to predict sale prices and dates for millions of homes daily and they evaluate their accuracy once a property is sold. Using live real estate data from Redfin, they are tracking trends in housing markets across the country. They are keeping tabs on which markets are heating up, where prices are cooling off, and how long it takes to sell a home.
In a space controlled by monopolies and gatekeepers, Nexplace seeks to provide a democratized network to evaluate home prices for everybody.
Miners
Miners will create their own models to forecast home prices and sale dates. They can either utilize data supplied by validators or access APIs to gather additional information for their model inference. Miners are responsible for providing the estimated sale date and predicted home price.
Validators
Validators supply data to miners from 50 markets, with plans to expand this number over time. They assess miners based on their accuracy in predicting home prices and sale dates.
Scoring Method for Home Price Prediction
The scoring system evaluates a miner’s prediction based on two main factors: the accuracy of the predicted home price and the accuracy of the predicted sale date. The final score combines these components with weighted contributions (86% for price accuracy and 14% for date accuracy). A perfect prediction yields a score of 100.
Awaiting Data
