{
  "fetched_at": "2026-10-20T01:53:37Z",
  "source_url": "https://www.kaggle.com/competitions/cameroon-air-quality-prediction/data",
  "pages": [
    {
      "name": "data-description",
      "content": "The dataset was extracted via the [open meteo API](https://open-meteo.com/). It contains daily measurements of weather and air quality data for ten cities in the country of Cameroon: Douala, Yaoundé, Bafoussam, Bertoua, Ebolowa, Bamenda, Buea, Ngaoundéré, Maroua, and Garoua.\n\n## Files\n\n*   **train.csv** - the training set\n*   **test.csv** - the test set\n*   **sample_submission.csv** - a sample submission file in the correct format\n\n\nFeatures description: \n\n| Feature | Description |\n|---------|-------------|\n| weather_code | The most severe weather condition on a given day (WMO code) |\n| temperature_2m_max | Maximum daily air temperature at 2 meters above ground |\n| temperature_2m_min | Minimum daily air temperature at 2 meters above ground |\n| temperature_2m_mean | The mean daily air temperature at 2 meters above ground |\n| apparent_temperature_max | Maximum daily apparent temperature |\n| apparent_temperature_min | Minimum daily apparent temperature |\n| apparent_temperature_mean | The mean daily apparent temperature |\n| daylight_duration | Number of seconds of daylight per day |\n| sunshine_duration | The number of seconds of sunshine per day (calculated by direct normalized irradiance exceeding 120 W/m², following the WMO definition) |\n| precipitation_sum | Sum of daily precipitation (including rain, showers and snowfall) |\n| rain_sum | Sum of daily rain |\n| precipitation_hours | The number of hours with rain |\n| wind_speed_10m_max | Maximum wind speed on a day |\n| wind_gusts_10m_max | Maximum wind gusts on a day |\n| wind_direction_10m_dominant | Dominant wind direction |\n| shortwave_radiation_sum | The sum of solar radiation on a given day in Megajoules |\n| et0_fao_evapotranspiration | Daily sum of ET₀ Reference Evapotranspiration of a well watered grass field |\n| city | Represents the city of the observation |\n| latitude | Represents the latitude of the location |\n| longitude |Represents the longitude of the location |\n| pm2_5 | Particulate matter with diameter smaller than 2.5 μm (PM₂.₅) close to surface (10 meter above ground) |\n| carbon_monoxide | Atmospheric carbon monoxide close to surface (10 meter above ground) |\n| nitrogen_dioxide | Atmospheric nitrogen dioxide close to surface (10 meter above ground) |\n| sulphur_dioxide | Atmospheric sulphur dioxide close to surface (10 meter above ground) |\n| ozone | Atmospheric ozone close to surface (10 meter above ground) |\n| aerosol_optical_depth | Aerosol optical depth at 550 nm of the entire atmosphere to indicate haze |\n| dust | Saharan dust particles close to surface level (10 meter above ground) |\n| uv_index | UV index considering clouds and clear sky |\n| uv_index_clear_sky | UV index considering clear sky |",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "Air pollution is a growing concern in rapidly developing African nations, with significant implications for public health and quality of life. Cameroon, like many countries in the region, faces increasing challenges in managing and predicting air quality, especially in its bustling urban centers.\n\nTraditional methods of air quality monitoring often require expensive equipment and extensive infrastructure, which can be prohibitive in resource-constrained environments. This competition proposes an innovative approach: leveraging readily available weather and environmental data to predict air quality levels, specifically PM 2.5 (Particulate Matter) concentrations.\n\nThe dataset provided encompasses daily measurements from ten major Cameroonian cities, including Douala, Yaoundé, and Bamenda. It includes a wide range of weather parameters such as temperature, precipitation, wind speed, and solar radiation, alongside air quality indicators. This comprehensive data offers a unique opportunity to uncover the intricate relationships between weather patterns and air quality in diverse urban settings across Cameroon.\n\nYour challenge is to develop a robust predictive model that can accurately forecast PM 2.5 levels based on these weather and environmental factors. A successful model could revolutionize air quality management in Cameroon by:\n\n1. Enabling timely public health warnings\n2. Informing urban planning and policy decisions\n3. Facilitating proactive measures to mitigate air pollution\n4. Enhancing our understanding of air quality dynamics in African urban environments\n\n\nBeyond the competition, the winning model will serve as the foundation for an End-to-End MLOps system for predicting Air Quality in Cameroon. \n\nThis system will showcase the practical application of machine learning in addressing real-world environmental challenges. \n\nIt will be developed as my final project for the [MLOps Zoomcamp](https://github.com/DataTalksClub/mlops-zoomcamp) organized by DataTalksClub, demonstrating the power of combining data science with robust engineering practices.\n\nYour contribution to this project extends beyond a mere competition entry. You'll be part of an initiative that bridges the gap between advanced data science techniques and pressing environmental needs in developing nations. This work has the potential to improve lives, inform policy, and set a precedent for similar applications across Africa and beyond.\n\n\n## Acknowledgements\n\nThe data used for this competition was extracted using the Open-Meteo API (https://open-meteo.com/). We extend our sincere gratitude to the Open-Meteo team for providing free access to high-quality weather and air quality data, which has been instrumental in making this project possible.\n\nOpen-Meteo's commitment to offering open and easily accessible environmental data has enabled us to compile a comprehensive dataset covering ten major cities in Cameroon. Their service has been crucial in bridging the gap between data availability and environmental research in regions where extensive local monitoring networks may not be present.\n\nWe would like to acknowledge the various data providers and meteorological organizations worldwide that contribute to the Open-Meteo database, ensuring its accuracy and reliability.\n\nSpecial thanks go to the developers and maintainers of the Open-Meteo API, whose work in aggregating and standardizing global weather and air quality data has significantly simplified the data collection process for this project.\n\nWe also express our appreciation to DataTalksClub for organizing the MLOps Zoomcamp, which has provided the framework and inspiration for developing this project into a full-fledged MLOps system. Their commitment to fostering practical data science skills has been invaluable.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "## Root Mean Squared Error (RMSE)\nSubmissions are scored on the root mean squared error. RMSE is defined as:\n\n$$\\text{RMSE} = \\sqrt{\\frac{1}{n} \\sum_{i=1}^{n} (y_i - \\hat{y}_i)^2}$$\n\nWhere \\\\(\\hat{y}_i\\\\) is the predicted value and \\\\(y_i\\\\) is the original value for each instance i.\n\n\n\n\n\n## Submission File\nFor each ID in the test set, you must predict the pm2_5 variable. The file should contain a header and have the following format:\n\n    id,pm2_5\n    4120,11.25\n    4121,10.9\n    4122,13.4\n    etc.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "Can you predict air quality levels in Cameroonian cities based on weather conditions and other environmental factors? This project aims to develop a predictive model that analyzes daily weather measurements and environmental data to forecast air quality, specifically PM 2.5 (Particulate Matter) levels. The dataset encompasses ten major cities across Cameroon: Douala, Yaoundé, Buea, Bamenda,  Ebolowa, Ngaoundéré, Bertoua, Maroua, Garoua, Bafoussam, providing a comprehensive view of air quality challenges in diverse urban environments.\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Timeline",
      "content": "- **Start date** - October 3rd, 2024\n- **Entry Deadline** - Same as the Final Submission Deadline\n- **Team Merger Deadline** - Same as the Final Submission Deadline\n- **Final Submission Deadline** - November 3rd, 2024\n\nAll deadlines are at 11:59 PM UTC+1 on the corresponding day unless otherwise noted. The competition organizers reserve the right to update the contest timeline if they deem it necessary.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}