AI-Powered Data Science — Diploma, Master & PG Diploma

Explain a pattern in existing data

Choose this focus when you enjoy investigating why a chart looks the way it does and which groups or periods contribute to a total.

Start with exploratory analysis, distributions and visualisation. State what each row represents, which comparison you made and what the data cannot explain.

Evaluate a prediction about unseen data

Choose this focus when you want to move from describing observations to testing whether a model is useful on examples it has not seen.

Connect preprocessing and statistics with the separate Machine Learning component. Compare a baseline and a model using the same evaluation rows.

Make a report someone else can use

Choose this focus when you want to translate analysis into a clear dashboard with named measures, useful filters and a concise explanation.

Connect SQL aggregation and Power BI relationships with a report question. Check how a filter changes the denominator before interpreting a percentage or trend.

Practice brief

A sales pattern with two possible explanations

Try this

Explore the synthetic sales CSV by month, product and region. Pick one visible change and compare units_sold with revenue rather than treating them as the same measure.

Keep this output

Two labelled charts and a short analysis note containing the observation, a plausible explanation and one question the dataset cannot resolve.

Check and explain

Check both the total and the product mix. A higher revenue total can reflect more units or a different mix; a chart alone does not prove the cause. Keep the synthetic teaching-data label on the report.

Syllabus connection

• Python
• Data Science using Python

Inspect the synthetic sales CSV

Practice brief

A forecast checked against later months

Try this

Aggregate the sales data by month. Use earlier months to explore a forecasting approach, then compare its estimates with months you kept aside.

Keep this output

A dated chart separating observed and predicted values, a table of errors, and a comparison with a simple baseline such as the previous month.

Check and explain

Keep later observations out of the preparation and fitting steps. Compare all methods over the same later period and discuss trend, seasonality and the short history before drawing conclusions.

Syllabus connection

• Data Science using Python
• Machine Learning

Use the sales teaching dataset

Practice brief

A dashboard with a reading guide

Try this

Build a Power BI report for a question such as “Which product and region contributed to this month’s units and revenue?” Use the published sales practice data.

Keep this output

A report page with a month selector, separate units and revenue measures, a product/region view and a brief guide explaining how to read the filters.

Check and explain

Reconcile an unfiltered total and one filtered total with a separate calculation. Check relationship behaviour and explain whether a displayed percentage is of all rows or only the selected group.

Syllabus connection

• SQL
• Power BI

Open the sales practice data