Data Analytics and Machine Learning for Weather Forecast

To enable students to leverage machine learning and deep learning to do weather prediction. They will learn to select relevant data from a big database and apply data analytics skills using machine learning predictive model to estimate temperature.

The school will organise a series of workshops. The workshops include theoretical session conducted in a classroom, introducing students to leverage different tools to perform data analytics. Basic level session may include interpreting and analysing data based on statistics and basic data visualisation tools in Spreadsheet like Pivot Table. Moderate level session may include more advanced data visualisation tools like Business Intelligence. For advanced level session, students will be taught to use the drag-and-drop tool and three main algorithms in Machine Learning, which doesn't require students to apply programming languages, to build, test, and deploy predictive analytics solutions on data. Students will then be required to display their findings and perform real-time analytics. Students from all sessions will be provided with big data (i.e. weathers related open data from Government such as temperature, humidity, winds and rainfall of selected automatic weather station by government or public observatory). Students are also allowed to source their own data to enrich their sample size.

After the workshop, the school will organise an intra-school competition for students. Students will be divided into teams and each team has to submit a total of 12 hourly temperature forecasts (from noon to evening) before deadline every day. Students' predictive results will be compared with the HKO instrument installed at around Hong Kong. Prizes will be awarded to teams with the most accurate model.

  • PC / laptop / tablet computer with Internet connection
  • Spreadsheet & business intelligence applications
Cloud Services
  • Machine learning service platform / APIs
  • Data provider APIs
Major Activities
  1. Theoretical session covering:
    1. Data collection from various sources (e.g. from government open data)
    2. Machine learning
      1. Three main algorithms applied in predictive analytics
      2. Big Data usage and application
      3. Computation
    3. Business intelligence
      1. Dashboard building
      2. Real-time analytical skills
    4. Spreadsheet
      1. Pivot Table
      2. Charts building
      3. Data sorting
      4. Data filtering

  2. Consulting session covering:
    1. Each group will be assigned a time slot to consult the teachers when they encounter difficulties in debugging problems.

  3. Share the challenges and lessons learnt
Learning Objective(s)
  • Enhance students' understanding of how digital tools or Artificial Intelligence can facilitate humans to make accurate decisions based on predictive model
  • Strengthen students' ability to interpret and analyse a large database
  • Nurture students' ability to apply the IT knowledge and skills learnt
  • Equip students with non-technical skills including collaboration skills, problem-solving skills and critical thinking skills
  • Theoretical session : 3-6 hours
  • Practical session : 3-6 hours
  • Medium
Target Level
  • Basic session : S1 - S2
  • Moderate level session : S3 - S4
  • Advanced session : S5 - S6
Target No. of Students
  • About 20 students per workshop
  • 4-5 students in a group