Time Series Analysis and Forecasting using Power BI

Name: Time Series Analysis and Forecasting using Power BI
Rating: 3.5 (25 reviews)

Learn how to use Power BI for time series exponential smoothing and handle errors using advanced Power Query techniques

Created byTobi Williams

Last updated 1/2021

English

English [Auto],

What you'll learn

Visualise time series data in Power BI
Apply and manipulate time series exponential smoothing forecast
Transform unstructured data into time series data
Understand time series theory, and the concepts of seasonal and cyclical data
Handle time series forecasting errors using advanced techniques in Power Query
Compare actual values versus forecast values

Course content

2 sections • 14 lectures • 1h 9m total length

Introduction to Time Series Data1:36
Explore time series data using Power BI by examining Google stock prices as a practical example. Learn how time-stamped observations form a time series data set.
What is Time Series Exponential Smoothing?2:16
Exponential smoothing forecasts future values by weighting past observations with exponentially decaying weights. It applies to univariate data, so recent observations carry more weight in predicting quantities like stock prices.
Forecasting Models in Power BI2:07
Use exponential smoothing to forecast seasonal and non-seasonal data, with Power BI automatically selecting the best model after analyzing historical data and weighting recent observations more.
What is Seasonality?1:58
Seasonality refers to regular variations within a time period, such as weekly, monthly, quarterly, or hourly patterns. Rainfall data and holiday-driven sales illustrate how demand fluctuates with festivities, impacting forecasting.
Cyclical Datasets0:54
Explore cyclical datasets that show peaks and dips not tied to fixed periods, not seasonal, driven by economic factors, and learn to identify these patterns in Power BI.
Check Your Knowledge

Generate Univariate Data for Analysis7:49
Analyze daily stock prices with time series and exponential smoothing in Power BI, generate a univariate data series, and explore short to long‑term forecasts.
Apply Forecast Line to Monthly Trend Analysis9:21
Apply a forecast line to monthly trend analysis in Power BI, using data aggregation to ensure evenly spaced dates and interpreting forecast horizons and confidence intervals.
Assess Model Accuracy7:32
Assess model accuracy in time series forecasting in Power BI by validating with a 95% confidence interval and hindcast tests that omit recent data; tune seasonality to avoid overfitting.
Advanced Error Handling - Create Date Table in Power Query11:01
Learn to handle missing data by building a complete date table in Power Query, inserting gaps, and calculating durations to support accurate time series analysis in Power BI.
Advanced Error Handling - Merge/Join Tables in Power Query3:40
Learn to perform a left join in Power Query, merging a date-backed table with a dataset to keep all dates and return the average value when matches occur.
Advanced Error Handling - Fill Blanks With Values5:33
Sort the data ascending and apply forward fill to missing values in Power BI, using prior data to maintain a consistent time series with a buffer table.
Apply Forecast Line to Daily Trend Analysis4:22
Apply a forecast line to daily trend analysis in Power BI using exponential smoothing for short-term time series forecasts, with 95 percent confidence intervals and seasonality considerations.
Questions - Time Series Exponential Smoothing7:40
Explore how closing prices drive time series aggregation from daily to monthly and yearly, and compare forecast with actual price using univariate exponential smoothing.
Further Reading - Describing the Forecasting Models in Power BI3:14
Predict NIKE stock prices

Requirements

A prior, basic understanding and usage of Power BI is recommended
Prior experience using Power Query or writing M scripts is useful but not required
An interest in stock trading is appreciated but not required
Have Power BI desktop and Microsoft Excel installed

Description

In this course, students will learn about the forecasting models available in Power BI. By understanding how time series exponential smoothing works, students will be able to manipulate the forecast line efficiently for daily, monthly, and yearly predictions of univariate data.

As part of the course, students will gain hands-on experience in advanced error handling techniques in Power Query and be able to tune parameters efficiently for cyclical and seasonal datasets.

What you’ll learn

Visualise time series data in Power BI
Apply and manipulate time series exponential smoothing forecast
Transform unstructured data into time series data
Understand time series theory, and the concepts of seasonal and cyclical data
Handle time series forecasting errors using advanced techniques in Power Query
Compare actual values versus forecast values

Are there any course requirements or prerequisites?

A prior, basic understanding and usage of Power BI is recommended
Prior experience using Power Query or writing M scripts is useful but not required
Interest in stock trading is appreciated but not required
Have Power BI desktop and Microsoft Excel installed

Who this course is for:

Business analysts interested in time series analysis
Finance professionals curious about Power BI and trend analysis
Python developers curious about Power BI and trend analysis
Power BI and Excel users interested in trend analysis
Business professionals curious about Power BI, forecasting, and time series analysis

Who this course is for:

Business analysts interested in time series analysis
Finance professionals curious about Power BI and trend analysis
Python developers curious about Power BI and trend analysis
Power BI and Excel users interested in trend analysis
Business professionals curious about Power BI, forecasting and time series analysis

Time Series Analysis and Forecasting using Power BI

What you'll learn

Explore related topics

Course content

Introduction to Time Series Exponential Smoothing5 lectures • 9min

Time Series Forecasting and Advanced Error Handling9 lectures • 1hr

Requirements

Description

Who this course is for: