Leveraging Business Intelligence (BI) unearths an organization’s strategies, trends, and emergent patterns based on data obtained from data sources already in use in the organization (Shauri 2023, 1). BI tools have proved to be the most go-to resource to help businesses harness the power of large data volumes and analytics, to make smarter, and data-driven decisions (Gurjar & Rathore 2013, 83).
One of such tools is Power BI that as a business analytics solution allows the user to analyse, visualize their data result, and share insights across the departments of the organization or embed the data results in the organization’s website (Arnold 2022). As data-analytics has continued to evolve, Power BI introduced Automated Machine Learning (AutoML) capabilities in 2019 available for Power BI Premium and Embedded dataflows environments (Philip 2019). This elevates the data-driven insights capabilities of the tool to include predictive capabilities that are key to forecasting while also saving time, and reducing costs as the model is generated automatically without manual efforts (Rawat 2023).
Machine learning-based prediction case study
This project utilized AutoML (Automated Machine Learning) integrated into the Power BI platform and Power BI Premium, was performed by 3Cloud Solutions, and was commissioned by the Georgia public school district. The project’s goal? To use predictive analytics and machine learning to predict students’ likelihood of graduating from high school based on a range of inputs, including attendance, behaviour, grades, and more. The school district could then proactively support and guide students who needed extra assistance.
AutoML streamlined the machine learning model creation process, making it straightforward. The first step was creating a dataset including the relevant data points that we believed could influence graduation. Then, we simply ran it through a wizard within Auto ML. The result? A fully trained model ready to make predictions.
But what exactly is AutoML in Power BI? AutoML is a wizard-driven machine learning interface within Power BI. It operates based on data flows, requiring a training entity and a prediction entity within the data flow. Once set up, a user selects the model they want to use, and AutoML automatically trains the model and delivers predictions. It’s a simple yet powerful pipeline that can leverage data stored in Azure Data Lake Gen2 and seamlessly imported it into SQL tables, as was with this case.
The outcome of this project thrilled the school district. They’re now actively using this machine learning model to predict high school graduation probabilities and provide vital support to students who might otherwise slip through the cracks. It’s a testament to the power of Azure and Power BI in making a real difference in education. (Turley 2020.)
Authors
Shauri Chelsea is a data sorcerer, conjuring insights, and predictions from the depths of raw information, and also a student at LAB University of Applied Sciences in a bachelor’s degree program in international business. With an interest in business intelligence, automation, machine learning, and Power BI, she transforms data into strategic gold.
Dr. Jukka Sirkiä is a senior lecturer at LAB University of Applied Sciences.
References
Arnold, J. 2022. Learning Microsoft Power BI. O’Reilly Media. Cited 21 Jul 2023. Available at https://learning.oreilly.com/library/view/learning-microsoft-power/9781098112837/ch01.html#power_bi_components
Gurjar, Y. S. & Rathore, V. S. 2013. Cloud business intelligence–is what business need today. International Journal of Recent Technology and Engineering, 1(6). Cited 6 May 2023. Available at https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=0e003c2cc7a9b7bc4c23e1bdc408d67f65a6b4b4
Mohamed_hassan. 2022. Palvelu, keinotekoinen, älykkyys, bot. Pixabay. Cited 4 Sep 2023. Available at https://pixabay.com/fi/vectors/palvelu-keinotekoinen-%C3%A4lykkyys-bot-6929022/
Philip, A. 2019. Announcing Automated Machine learning in Power BI general availability. Microsoft Power BI Blog. Cited 2 Sep 2023. Available at https://powerbi.microsoft.com/en-in/blog/announcing-automated-machine-learning-in-power-bi-general-availability/
Rawat, M. S. 2023. Machine Learning with Power BI: Unlocking the Potential. Cynotech. Cited 3 Sep 2023. Available at https://cynoteck.com/blog-post/machine-learning-with-power-bi/
Shauri, C. 2023. Real-time business intelligence : data-driven decisions for sales stakeholders. Bachelor’s thesis. LAB University of Applied Sciences, international business. Cited 2 Sep 2023. Available at https://urn.fi/URN:NBN:fi:amk-2023090125204
Turley, P. 2020. A Success Story Using Auto ML In Power BI Premium. Cited 3 Sep 2023. Available at https://3cloudsolutions.com/resources/a-success-story-using-auto-ml-in-power-bi-premium/
