Quick Stats

Quick Stats

    You are here

    Machine Learning Unlocks Power of the Cloud, Big Data

    Leveraging predictive analytics solutions

    By ShiSh Shridhar, Microsoft

    Today retailers have access to a variety of data sources that include social media, open data and data generated by the Internet of Things. Combining proprietary data with public and purchased data provides retail businesses with valuable insights into customer purchase patterns and demand. Machine learning -– the ability to apply historical data to a problem and use that model to successfully predict future consumer behavior or trends -– enables retailers to use this vast trove to program systems that can make predictions and inferences based on data patterns.

    Here’s a retail example: an online customer who is browsing a store for a camera places the item in his shopping cart and is then recommended to buy an accessory -- a tripod. This is a simplistic example but machine learning can empower retailers to go further by uncovering unexpected buying patterns, based on unforeseen relationships between different customers and products. Machine learning algorithms are an evolution beyond normal algorithms. They make your programs "smarter" by allowing them to automatically learn from the data you provide. With the availability of data from various sources, retailers can use customers' online click history, purchase history, social networks and demographic information to refine the recommendations.

    Machine learning capabilities already touch many facets of our lives today -- from search engines to online product recommendations, credit card fraud prevention systems and mobile personal phone assistants like Cortana. But we’ve barely scratched the surface of its potential to change the world. We need to make machine learning more accessible –- to every enterprise and, over time, every one.

    Machine Learning Algorithms

    In the era of ambient intelligence, machine learning makes scenarios like these a reality:

    A retailer is able to predict the right recommendations from a customer’s purchase history, current inventory, the customer’s current location based on mobile data, social media as well as external weather information. The retailer analyzes all of this in seconds so it can send the customer a voucher for a BBQ cleaner to their phone –- but only if they own a barbeque, the weather is nice and they are currently within a specific distance of a retail store that has the BBQ cleaner in stock. The prediction of what offer to make, what channel to use for the offer and when to make the offer can be made using machine learning algorithms.

    Today, machine learning is usually self-managed and on-premises, requiring complex systems, training and expertise of data scientists. However, data scientists are in short supply, commercial software licenses can be expensive and popular programming languages for statistical computing have a steep learning curve. Even if a business could overcome these hurdles, deploying new machine learning models in production systems often requires months of engineering investment. Scaling, managing and monitoring these production systems requires the capabilities of a sophisticated engineering organization, which few retailers have today.

    To help overcome these challenges, Microsoft recently announced Azure Machine Learning, a fully-managed cloud service for building predictive analytics solutions, to overcome the challenges most businesses have in deploying and using machine learning. The visual model development environment enables organizations without resident data experts to derive value from Azure ML. Azure ML has a large set of pre-defined algorithms for analyzing datasets, which users can drag and drop to select and combine together to develop their model. With Azure ML, the cloud eliminates the need to procure and setup in-house servers and provides the ability to scale on demand. In mere hours, with Azure ML, partners can build data-driven applications to predict, forecast and change future outcomes -– a process that previously took weeks and months.

    For example, Microsoft partner MAX451 is using the service to help a large retail customer determine what products a consumer is most likely to purchase next, based on ecommerce data as well as brick and mortar store data. Combined with Microsoft’s data platform, Azure Machine Learning will help partners like MAX451 create entirely new solutions that bring together big data insights, the Internet of Things and predictive analytics.

    A public preview of the Azure Machine Learning service will be available this month. To learn more about how machine learning can transform our world, check out this video.


    By ShiSh Shridhar, Microsoft
    • About ShiSh Shridhar ShiSh Shridhar is director, big data solutions for retail, Microsoft

    Related Content

    Related Content