# SAS Machine Learning Model Reveals the World‘s Best Places to Live for 2026 - Canonical: https://33rdsquare.com/sas-built-machine-learning-model-rank-the-best-places-world/ - Published: 2024-09-03 - Author: Jordan Brown - Categories: [Artificial Intelligence & Machine Learning & ChatGPT](https://33rdsquare.com/category/tech/ai/) --- Have you ever wondered what the best places in the world to live are? Where does your city rank? What if there was an objective way to determine the top cities for quality of life? The analytics company SAS has done just that by building a sophisticated machine learning model to rank the best places to live worldwide. SAS is a leading analytics platform that empowers organizations to transform data into intelligence. By leveraging the power of machine learning, SAS was able to analyze over 5 million data points to determine the most liveable cities globally. Let‘s dive into how they built this impressive model and reveal the best places in the world to live in 2024 according to SAS. ## How Machine Learning Enables Data-Driven Rankings at Scale Machine learning is a branch of artificial intelligence that allows systems to automatically learn and improve from experience without being explicitly programmed. Machine learning algorithms build models based on sample data in order to make predictions or decisions on new data. This ability to learn patterns from large, complex datasets makes machine learning an ideal tool for ranking and recommendation tasks like determining the best cities to live in. With traditional methods, it would be impossible for humans to manually analyze millions of data points across hundreds of factors and locations. Machine learning excels at finding relationships, correlations and insights in massive datasets that humans would almost certainly overlook. By training a machine learning model on global data about quality of life factors, SAS was able to uncover the cities that offer the best combinations of those factors to rank them in a data-driven, objective manner. ## Inside the SAS Machine Learning Model To build the model, SAS collected an extensive dataset from a wide variety of sources, including: - Public data from UNESCO and the World Tourism Organization - Social media posts and check-ins - Travel review websites like TripAdvisor - Geotagged data on climate, pollution levels, green spaces, etc. In total, the dataset contained over 5 million data points, far more than any individual human could sort through. The researchers used machine learning techniques to process and analyze this huge volume of information. Initially, hundreds of raw features were considered, but through a process called feature selection, the team narrowed it down to the eight most predictive factors: 1. Cost of living 2. Infrastructure and safety 3. Healthcare accessibility and quality 4. Education and employment opportunities 5. Family-friendliness 6. Culture and leisure activities 7. Shopping and dining options 8. Environmental factors like air quality, green space, etc. The data scientists then transformed the raw data into informative features for the model. For example, rather than using the raw count of restaurants, they calculated the restaurants per capita. Rather than total green space, they calculated green space as a percentage of total city area. The cleaned, transformed dataset was split into a training set and test set. The training set was used to train the machine learning model, while the test set was used to evaluate how well the model performs on new, unseen data. Several machine learning algorithms were evaluated, but a random forest model was ultimately selected for its strong performance and interpretability. Random forests combine multiple decision trees to make predictions. They are robust against overfitting and can rank the importance of input features. Hyperparameter optimization was used to tune the model for best performance. Hyperparameters are configuration settings that are not learned from the data but set before training. The final random forest model was then used to predict quality of life scores for each factor for cities around the globe. Those scores were weighted based on global survey data about the relative importance of each factor. The weighted scores were totaled to produce the overall global rankings. ## The Best Cities to Live in 2024 According to SAS So which places topped the list? According to the SAS model, the 10 best cities to live in the world for 2024 are: 1. Zurich, Switzerland 2. Calgary, Canada 3. Stockholm, Sweden 4. Helsinki, Finland 5. San Francisco, USA 6. Auckland, New Zealand 7. Perth, Australia 8. Vienna, Austria 9. Tokyo, Japan 10. Melbourne, Australia Zurich landed the #1 spot with its high marks for safety, infrastructure, education, and environmental quality. Residents enjoy an extremely high quality of life, with 96% reporting they are satisfied or very satisfied. The city‘s extensive train, bus and tram system makes it easy to get around. Zurich also boasts over 50 museums, more than 100 art galleries, and world-class dining. North American cities Calgary and San Francisco made the top 10 for their combination of employment opportunities, leisure activities, cultural diversity, and family-friendliness, even with a higher cost of living. Calgary has a low population density with over 10,000 hectares of parkland. San Francisco is home to some of the world‘s most innovative tech companies and a thriving food scene. ![Chart showing cost of living vs quality of life rating for top 10 cities](https://33rdsquare.com/top10_col_vs_qol.png) A cluster of Scandinavian and Nordic cities ranked highly, which is no surprise given their reputations for progressive city planning focused on sustainability, equality, and wellbeing. Stockholm, Helsinki, and Vienna offer residents an outstanding balance across all eight factors used in the model. They have efficient public transit, excellent education and healthcare, ample green space, and a cultural focus on work-life balance. Australia and New Zealand are represented by Perth, Melbourne, and Auckland, which each scored very well on environmental quality and outdoor activities. Over 80% of residents in these cities live within walking distance of a park. Tokyo is the only Asian city in the top 10, earning its spot for safety, dining, and efficient infrastructure. It has one of the lowest crime rates of any major global city. Noticeably absent are some of the world‘s largest and most famous cities like New York, London, and Paris. While they scored very well in certain categories like culture and dining, their overall rankings were hurt by lower scores on cost of living, environmental factors, and family-friendliness. Mexico City and Sao Paolo ranked highly for affordability and leisure but missed the top 10 due to lower healthcare and safety scores. ## Benefits and Applications of AI-Driven City Rankings The SAS model demonstrates the power of machine learning to provide data-driven, objective insights that would be impossible for humans to match. No individual could analyze 5 million data points across hundreds of cities worldwide. By removing human biases and limitations, this model offers a more accurate way to assess quality of life at a global level. It doesn‘t rely on perceptions, opinions, or anecdotal evidence, but rather cold, hard data. And a lot of it. The ability to process and find patterns in huge datasets is one of the key benefits of AI. With a machine learning approach, SAS was able to create a much more comprehensive evaluation of what makes a city great to live in compared to traditional methods. Some additional benefits and applications of AI-driven city rankings include: - **Personalization**: The model could be adapted to create custom rankings for individuals based on their unique preferences and needs. For example, a family with young children would have different priorities than a single young professional. - **Simulation**: Machine learning models can be used to simulate what would happen to a city‘s ranking if certain factors changed. This could help urban planners test the impact of policy proposals, infrastructure investments, etc. - **Real-Time Monitoring**: By integrating real-time data streams, an ML model could continuously update rankings and provide early warning of any declines in quality of life due to increased pollution, cost of living spikes, etc. - **Optimization**: AI techniques like reinforcement learning could suggest actions for cities to take to improve their rankings and quality of life for residents in the most efficient manner. ## Limitations and Future Enhancements As with any machine learning model, the SAS global ranking system does have some limitations. While the dataset is extensive, it may not capture every single factor that influences quality of life. There could also be biases in the training data that influence the results. The model also represents a single point in time. Quality of life is dynamic and rankings could change substantially due to major events, policy changes, etc. Retraining the model frequently on refreshed data would help provide a more up-to-date assessment. There are many opportunities to enhance and refine the model over time. More granular geographical data could enable neighborhood-level rather than just city-level rankings. Applying more advanced techniques like deep learning could uncover even more nuanced insights. Here are a few additional ideas for future enhancements: - **Sentiment Analysis**: Incorporate sentiment analysis of social media posts, news articles, etc. to gauge public perception of quality of life factors and how they change over time. - **Anomaly Detection**: Use anomaly detection methods to automatically identify and flag any drastic changes in feature values that could signal major quality of life disruptions. - **Explainable AI**: While the random forest model provides some insight into feature importance, applying explainable AI techniques could shed even more light on why the model makes the predictions it does. - **Multi-Objective Optimization**: Investigate multi-objective optimization methods to find the best possible trade-offs between quality of life factors, since it‘s often impossible to maximize all of them simultaneously. ## The Endless Possibilities of AI for Enhancing Quality of Life The SAS machine learning model for ranking the world‘s best places to live showcases the truly revolutionary potential of AI and big data. It gives us the ability to make more informed, evidence-backed decisions about where to live and travel. It provides an objective starting point for exploring new places. And it pushes cities to continue improving quality of life for residents. As the authors of the classic AI textbook "Artificial Intelligence: A Modern Approach" put it, "AI is the science and engineering of making intelligent machines." The SAS model is a prime example of using AI to extract intelligent insights from vast amounts of data. A 2019 study by McKinsey estimated that AI has the potential to create up to $15.4 trillion in annual global economic value by 2030. While much of that will be in business settings, AI also has immense potential to improve quality of life for individuals around the world. What unique factors make somewhere a great place to live for you? What matters most in your personal location decisions? With machine learning, we can help uncover the places and features that will maximize quality of life for each person‘s unique situation. We can simulate the impacts of potential changes. And we can continually monitor for any issues. The possibilities for AI to enhance human life across all facets – where we live, work, learn, play – are truly endless. By continuing to advance machine learning capabilities and apply them in data-driven ways, we can make smarter decisions to improve quality of life for all. --- Source: [SAS Machine Learning Model Reveals the World‘s Best Places to Live for 2026](https://33rdsquare.com/sas-built-machine-learning-model-rank-the-best-places-world/)