Dr. Lei Tang and Dr. Xin Xu will talk about how they apply machine learning with small data sets in sales management and forecast.
The recent successes of machine learning and deep learning can be largely attributed to three factors: emergence of abundant data, development of innovative algorithms, and availability of machine learning tools and computing resources. Unfortunately, not all application spaces provide data sets large enough to be used in the usual or obvious ways. In this talk, Lei and Xin focus on one specific domain, enterprise sales, where data is often limited in volume, always noisy, and constantly evolving. They describe how machine learning, and in particular deep learning, can help, and how we address the data challenges described. They specifically discuss how to select model architectures appropriate for these limited data situations, for example, how deep our networks should be. By sifting through sales records and associated sales activities Lei and Xin enable identification of at-risk opportunities as well as project and estimated the time required to close each deal. This, in turn, contributes to the generation of a reliable business forecast for sales managers and executives. Lessons and findings learned through the process is shared.
Dr. Lei Tang is the Chief Data Scientist at Clari Inc., a startup backed by Sequoia Capital and Bain Capital ventures, focusing on predictive analytics for sales execution and forecasting. Lei received his Ph.D. in computer science from Arizona State University in 2010, and B.S. from Fudan University, China. He is passionate about reshaping variety of businesses, driving business growth and decision through data science and machine learning. From 2012-2014, Lei was the lead data scientist at Demand Generation of @WalmartLabs, where he worked closely with marketing team to drive traffic to site, impacting hundreds of revenue each year. Before that, Lei had 2-year stint at advertising sciences in Yahoo! Labs, working on targeting, user profiling/segmentation by mining user behavioral, social and content information. Lei has co-authored one book on “community detection and mining in social media” (top-download in the corresponding data mining lecture series), held 4 patents, published over 30 papers at top-notch conferences and journals on data mining/machine learning, with over 4000 citations.
Dr. Xin Xu is currently working as a data scientist in Clari. Before this, She received her Ph.D degree in Computer Engineering from North Carolina State University in 2015. She also did summer intern in Bell Labs and Akamai Technology in 2014 and 2015 respectively. Her current research interest mainly focuses on applying data mining, machine learning and advanced analytics to solve practical problems in sales domain.