Thirteen teams representing 70 students from Statistics and Data Science’s Master of Professor Studies program in Applied Statistics presented their semester-end research projects at a poster session on Monday, May 8.
Students applied statistical methods to projects ranging from e-scooter driving patterns in Washington D.C. and food security in Malawi, to clothing design and battery-life prediction.
Held every May, this poster session showcases student research for the MPS program’s linchpin project course, Applied Statistics MPS Data Analysis Project (STSCI 5999), which challenges students to work alongside client companies and apply statistical methods to solve real-world problems.
Faculty judges selected the projects below based on their scientific rigor, poster design, and appearance. Students were also rated on their ability to communicate their research goals and methods, provide informative and well-organized posters, and deliver engaging and concise presentations.
"Forecasting Wheat Production in Morocco"
Team members: Yudong Fang, Mingyi Hou, Simin Le, Yunhao Li, Jingyan Liu, and Hanyue Yao; under the guidance of advisor Ahmed El Alaoui, assistant professor of statistics and data science
Client: Gro Intelligence
This team explored the potential to predict annual winter wheat production in Morocco – a major wheat producer in Africa – using historical data on precipitation, land temperature, soil moisture, and other factors. They found that two prediction models, “sliding window” and Long Short-Term Memory (LSTM), appear to be the best options for informing agricultural decisions concerning wheat. This team impressed judges with its scientific excellence, innovative approach, and clear presentation style.
"Analyzing Lower Torso Curves for Computational Apparel Design and Modeling"
Team Members: Wenxuan Hong, Aristotle Kolefas, Diane Xu, Zongjie Yin, and Fanjun Zeng; under the guidance of advisor Sreyoshi Das, assistant professor of practice in the Department of Statistics and Data Science.
For women, finding well-fitting pants can be challenging. This team investigated whether data science methods could be used to inform the design of women's pants based on individual body shapes captured by a 3D body scanner. The team’s findings suggest that data-driven features extracted from curves can help guide apparel design for groups of women with characteristically different body shapes. Judges noted the poster’s visually appealing design and informative content.
The STSCI 5999 course provides an excellent opportunity for students to develop their research and presentation skills and to make meaningful contributions to the field of statistics. Learn more about the MPS program in Applied Statistics on the Statistics and Data Science website.