Presentation is loading. Please wait.

Presentation is loading. Please wait.

Project #6: Processing data from a visual IoT sensor of Fluid/Structure interactions with Machine Learning REU Students: Alexis Downing and Pete Orkweha.

Similar presentations


Presentation on theme: "Project #6: Processing data from a visual IoT sensor of Fluid/Structure interactions with Machine Learning REU Students: Alexis Downing and Pete Orkweha."— Presentation transcript:

1 Project #6: Processing data from a visual IoT sensor of Fluid/Structure interactions with Machine Learning REU Students: Alexis Downing and Pete Orkweha Graduate mentor(s):Safa Bacanli and Sharare Zehtabian Faculty mentor(s): Andrew Dickerson and Damla Turgut Week #5 (June 24 – June 28, 2019) Accomplishments: Continued the literature review Experimentation and implementation of the Gradient Boosting Regressor with Ridge Regression. The specified ridge regression used had cross validation. These algorithms was used to try to reduce the root mean squared error (RSME)and relative error(RE). Contact time is calculated by the frame delta time multiplied by the total number of frames. The starting frame occurs when, the water droplet is one frame before touching the fiber. The ending frame occurs when the water droplet is visually, fully separated from the fiber and/or it is not connected to the remaining droplet surrounding the fiber. Problem & Solutions Maximum deflection verses drop velocity was a recurring problem. Through physics, the force applied to the fiber it self results in a lower maximum deflection. Ensemble learning algorithms were applied to find the best scores for the root mean squared error, relative error, and R^2. Plans for next week: Investigate how much energy is transferred to the fiber and how contact time affects the maximum deflection. Record more data for better accuracy and precision.

2 Starting frame & Comparison of two ending frames
Fully separated (ending) Starting frame Not Fully separated(ending)


Download ppt "Project #6: Processing data from a visual IoT sensor of Fluid/Structure interactions with Machine Learning REU Students: Alexis Downing and Pete Orkweha."

Similar presentations


Ads by Google