{"id":3400,"date":"2023-06-29T12:07:34","date_gmt":"2023-06-29T12:07:34","guid":{"rendered":"https:\/\/neuraldesigner.com\/blog\/madrid-air-forecasting\/"},"modified":"2025-09-15T16:03:21","modified_gmt":"2025-09-15T14:03:21","slug":"madrid-air-forecasting","status":"publish","type":"blog","link":"https:\/\/www.neuraldesigner.com\/blog\/madrid-air-forecasting\/","title":{"rendered":"Forecast air pollution in a city using machine learning"},"content":{"rendered":"

Air pollution is one of the significant problems the world faces.<\/p>\n

Having a way to monitor these levels, allowing for informed decisions, can be crucial for public administrations.<\/p>\n

We want to predict the levels of atmospheric pollutants for the following week in Madrid.<\/p>\n

We solve this example with Neural Designer<\/a>. To follow it step by step, you can use the free trial<\/a>.<\/p>\n

\n

Contents<\/h2>\n
    \n
  1. Application type<\/a>.<\/li>\n
  2. Data set<\/a>.<\/li>\n
  3. Neural network<\/a>.<\/li>\n
  4. Training strategy<\/a>.<\/li>\n
  5. Model selection<\/a>.<\/li>\n
  6. Testing analysis<\/a>.<\/li>\n
  7. Model deployment<\/a>.<\/li>\n<\/ol>\n<\/section>\n
    \n

    1. Application type<\/h2>\n

    This is a forecasting<\/a> project since the variables to be predicted are the values of five pollutants for the next week.<\/p>\n

    The goal here is to obtain an accurate prediction for each of them.<\/p>\n<\/section>\n

    \n

    2. Data set<\/h2>\n

    The file madrid-forecasting.csv<\/a> contains 2,959 samples, each with 15 inputs.<\/p>\n

    Variables<\/h3>\n

    We transform the dataset into a time series with lags and steps ahead. The output variables are 35, one for each pollutant, for 7 days of the week. The following list summarizes the variables’ information:<\/p>\n

    Regarding time attributes:<\/p>\n