• National Tsing Hua University logo

Business Analytics Using Forecasting

Learn a scientific and practical approach for creating and evaluating forecasting solutions.

Business Analytics Using Forecasting
  • Duration6 weeks
  • Weekly study3 hours
  • CertificatesAvailable

Why join the course?

Companies, governments and other organizations now collect and analyze huge amounts of data about suppliers, clients, employees, citizens, transactions, and much more. There are a number of ways organizations can use this data. Business analytics uses this data to make better decisions and forecasting is an arm of this predictive analytics. Forecasting especially can provide a powerful toolkit for analyzing time series data.

Learn about forecasting in a wider context

Quantitative forecasting uses statistical and data mining methods to generating numerical forecasts, an important component of decision making across many business functions, including economic forecasting, workload projections, sales forecasts, and power and transportation demand. Today’s big data forecasting can include forecasting many series on a frequent basis, such as daily demand of thousands of products at retail chains, hourly statistics of wind turbines, minute-by-minute web traffic, and call volume to call centers.

Forecasting can also be combined with statistical monitoring methods for purposes of anomaly detection – for example, public health organizations collect and monitor clinical and other data for detecting disease outbreaks. Forecasting is also often combined with simulation for purposes of scenario building. On this course we’ll have a look at some of these uses in more depth as well as examining the processes that these different industries use.

Understand the forecasting process

This course focuses on forecasting time series, where past and present values are used to forecast future values of a series of interest. The course covers issues relating to different steps of the forecasting process, from goal definition, through data visualization, modeling, and performance evaluation to model deployment.

In this course you will:

  • Learn how to define a forecasting task and workflow
  • Understand how to evaluate forecasting performance
  • Apply and be familiar with popular forecasting methods
  • Explore, identify and model different types of patterns in time series
  • Be able to implement a forecasting process in practice
  • Download video: standard or HD

    Skip to 0 minutes and 10 seconds Hello I’m Galit Shmueli, a professor at National Tsing Hua University’s Institute of Service Science I design and teach courses on Business Analytics to help students understand not just how to crunch data but why to crunch data. People, societies, products and processes generate lots of data. Data science and business analytics have evolved as more and more data becomes available at higher resolution and at higher frequency. Quantitative forecasting is the science of using time series data for generating forecasts. In other words, extrapolating a series of measurements into the future. Forecasting plays an essential role in decision making in almost any environment you can imagine where data are collected over time. Here are just a few examples.

    Skip to 1 minute and 2 seconds Forecasting the number of hourly or daily customers at a restaurant, bank, or other service provider can help with staffing decisions, inventory management, cash flow planning, service decisions, and even space design and choice of technologies to use. Forecasting is essential also in transportation. Think about bicycle sharing systems. Forecasting the demand for bicycles at each station at different times during the day can help with smarter distribution of bikes expansion of bike stations and planning new locations and bike routes. It can even help with maintenance scheduling. With today’s Internet of Things (IoT) automated sensor data can help forecast usage of offices, restrooms, vending machines, and other facilities. Such forecasts can help with staffing, replenishment, and other service operations.

    Skip to 1 minute and 59 seconds It can also support environmentally minded decision making and design. In this course, we’ll focus on how to create a forecasting solution. You’ll learn about different popular forecasting methods and algorithms but we’re also going to focus on the entire forecasting process. This means we’re gonna look at how to define a forecasting problem, how toevaluate the performance of a forecasting method, and, importantly, how to tie the forecasting analytics with the business problem. This course is for you if you’d like to extend your predictive analytics capabilities to the forecasting domain and time-series domain or if you have to generate forecast for your job. Maybe you’d like to become more knowledgeable consumer of forecasts.

    Skip to 2 minutes and 50 seconds Or maybe you’re just curious to learn scientific and practical approach to forecasting time series. The learning in this course is based on weekly activities In each week, you’ll be expected to read a chapter in the textbook, to watch a few short videos, and to work on an activity or analysis related to forecasting. Finally you’ll engage in an online forum. This combination is designed to help you learn about forecasting in an active and engaging way. I hope you now have a better idea about what this course is about and what you might learn from it. Welcome to the community of learners on time series forecasting or as we say here 歡迎光臨(Huānyíng guānglín)!

    What topics will you cover?

    The course covered issues relating to different steps of the forecasting process:

    • goal definition
    • data visualization
    • modeling
    • performance evaluation
    • model deployment

    What will you achieve?

    By the end of the course, you‘ll be able to...

    • Describe business challenges and opportunities that call for forecasting
    • Evaluate performance of a forecasting solution
    • Apply and be familiar with popular forecasting methods
    • Explore, identify and model different types of patterns in time series
    • Develop a forecasting solution using forecasting methods

    Who is the course for?

  • Familiarity with basic statistical methods including linear regression.
  • Basic knowledge of Excel and R software.
  • What do people say about this course?

    A great introduction to forecasting. It cleared up some problems that I was finding where I was trying to overfit to Holt-Winters and not getting good results. The training and validation periods were a bit of an eye-opener. I got a bit lost on autoregression techniques particularly ARMA and ARIMA, but I will go through those modules again, to try and make sense of it. I also purchased the book so this should also help. The R programming language is much more powerful than I had thought and the step by step process will help. There are some good tutorials on how to use it on the web that has been helpful. Thank you for putting this course online.

    Jonty Pearce

    Who will you learn with?

    Distinguished Professor @ NTHU, Taiwan. Pioneered business analytics courses at U of Maryland, Indian School of Business, NTHU & Statistics.com. Read about her research & textbooks at galitshmueli.com

    Who developed the course?

    National Tsing Hua University

    National Tsing Hua University in Taiwan consistently ranks as one of the premier universities in East Asia.

    Learner reviews

    See all FutureLearn Reviews

    Buy a personalised, digital and printed certificate and transcript

    You can buy a Certificate of Achievement for this course — a personalised certificate and transcript in both digital and printed formats, to prove what you’ve learnt. A Statement of Participation is also available for this course.

    Certificate of Achievement + transcript


    Statement of Participation