SAS’s Machine Learning Model Ranks the Best Places to Live in the World

Overview

• SAS researchers have developed an algorithm that ranks the best places to live in the world
• The AI model was trained on more than 5 million data points sourced from popular websites
• 8 key variables were taken for the final model

Introduction

The debate on which place(s) is the best in the world has been raging on since times immemorial. But thanks to SAS and machine learning, we might finally have the answer.

Researchers at SAS claim to have created an artificial intelligence program that ranks the best places to live in the world using publicly available data.

The model was developed using more than 5 million data points sourced from various social media platforms, travel review websites (like TripAdvisor) websites, public statistics from UNESCO and WTO and also used geodata. The final data that was analysed and trained on included the weather, job prospects, pollution, public transport, economy, healthcare and green locations.

Using a variable reduction technique, the key criteria for the final algorithm were cut down to eight, namely:

• Expenses
• Infrastructure and safety
• Healthcare
• Education and employment
• Attractiveness to families
• Culture
• Shopping and restaurants
• Environment

The results of the above criteria were then compared to quality of life indicators, like the price of local fruits and groceries, the number of walkable pavements, the number of trees in the area, the width of footpaths and even the likelihood of traffic jams.

The top 5 places to live, according to the algorithm, are:

1. West Perth, Australia
2. Feijenoord, Rotterdam, Netherlands
3. New York, NY, United States
4. Sandy Bay, Australia
5.  Hebden Bridge, United Kingdom

Our take on this

While debating and analysing the best places to live or travel in the world, our unconscious bias inevitably creeps into the equation. What this research from SAS does is it takes out that human element and by analysing more data than ever before, it has truly trumped every argument.

Once more details are revealed about the research and depending on it’s commercial release, it could also help house buyers in comparing and buying real estate. Let’s hope we see the source code behind this algorithm soon so we can try it out as well!

Subscribe to AVBytes here to get regular data science, machine learning and AI updates in your inbox!

Also, go ahead and participate in our Hackathons, including the DataHack Premier League and Lord of the Machines!

Pranav Dar

Senior Editor at Analytics Vidhya.Data visualization practitioner who loves reading and delving deeper into the data science and machine learning arts. Always looking for new ways to improve processes using ML and AI.