University of California, San Diego
• Implemented Particle Filtering to interpolate and predict Unevenly Spaced Time Series data points and convert the data into evenly spaced time series for data analysis on low power devices.
• The particle filter algorithm was developed using C++ and ran on Raspberry Pi 2 and A+ over meteorological data and obtain real time interpolation and prediction results.
• Measured accuracy and power consumption of particle filtering to verify how suitable they are for low power embedded devices.
Paper can be found here:http://bit.ly/2a2KZvN
