6. Data Applications

In this topic, we will focus on derivative applications using Civil IoT Taiwan open data. We strengthen the value and application services of Civil IoT Taiwan’s open data by importing other library packages and analysis algorithms. The units we expect to develop include:

• 6.1. Machine Learning Preliminaries

  We use air quality and water level data, combined with weather observations, using machine learning for data classification and data grouping. We demonstrate the standard process of machine learning and introduce how to further predict data through data classification and how to further explore data through data grouping.

• 6.2. Anomaly Detection

  We use air quality data to demonstrate the anomaly detection framework commonly used in Taiwan's micro air quality sensing data. We learn by doing, step by step, from data preparation and feature extraction to data analysis, statistics, and induction. The readers will experience how to gradually achieve advanced and practical data application services by superimposing basic data analysis methods.

• 6.3. Joint Data Calibration

  We use air quality category data of the Civil IoT Taiwan project to demonstrate the dynamic calibration algorithm for Taiwanese micro air quality sensors and official monitoring stations. In a learning-by-doing way, from data preparation, feature extraction, to machine learning, data analysis, statistics and induction, the principle and implementation process of the multi-source sensor dynamic calibration model algorithm are reproduced step by step, allowing readers to experience how to gradually realize by superimposing basic data analysis and machine learning steps to achieve advanced and practical data application services.