Student Research Reports
A Hutchinson-Based Approach to Mosquito Modeling: Predicting Mosquito Threat with Machine Learning and Edge Computing
Country:United States of America
Student(s):Avi Bagchi, Govind Gnanakumar, Shyam Polineni, Sujay Rasamsetti, Om Shastri, Gianna Yan, and Spencer Burke
Grade Level:Secondary School (grades 9-12, ages 14-18)
GLOBE Teacher:Cassie Soeffing
Contributors:Dr. Rusty Low, scientist, IGES
Peder Nelson, scientist, OSU
Dr. Erika Podest, scientist, NASA JPL
Report Type(s):International Virtual Science Symposium Report, Mission Mosquito Report
Protocols:Land Cover Classification, Earth As a System, Mosquitoes
Presentation Video:
View Video
Presentation Poster:
View Document
Optional Badges:I am a Data Scientist, I am an Engineer, I make an Impact
Language(s):English
Date Submitted:02/10/2022
Species distribution models (SDMs) that use climate variables to make binary predictions are effective tools for niche prediction in current and future climate scenarios. In this study, we defined a Hutchinson hypervolume with temperature, humidity, air pressure, precipitation, and cloud cover climate vectors collected from the National Oceanic and Atmospheric Administration that we matched to mosquito presence and absence points extracted from GLOBE Observer and the National Ecological Observatory Network. We then trained an 86% accurate Random Forest model that operates on binary classification. Given a location and date input, the model produces a threat level based on the number of decision trees that vote for a presence label. The feature importance chart and regression shows a positive, linear correlation between humidity and mosquito threat, as well as between temperature and threat below a threshold of 28 C. We then visualized the results in an interactive ArcGIS Dashboard. In accordance with the aforementioned statistical analysis, we found high threat clusters in warm, humid regions. We also built a device leveraging GPS smartphone technology and the IoT to collect and analyze data on the edge, allowing for users to obtain a real time threat level in remote areas without cloud connectivity.
Keywords: mosquito prediction, machine learning, edge computing, citizen science