To build a Machine learning model that is able to predict the demand of a newly launched product with features not currently present in any other currently available product. Since new model had features never seen before the dataset set created comprised of products which nearly resembled the product/feature. K-means and K-modes clustering was applied in a pipeline in order to get the cluster from the dataset that matched most with the product. The prices of these products over the last 5 years was then passed to a Bass-Diffusion Model for predicting the demand.
2018-9 - 2018-9
Nature inspired algorithms such as Bat Algorithm, Firefly algorithm studied and implemented for integration in image enhancement
2018-8 - 2018-12
Wrote MATLAB scripts for Extended Kalman Filter, Unscented Kalman Filter for a non-linear motion and measurement model.
2018-5 - 2018-7
Drone Localization Using Whycon Markers
Whycon Markers were used in order to Localize a Drone and provide aerial stability using PID Controller.
2017-9 - 2018-3
Surveillance System Using Computer Vision and Internet of Things Manipal
Project included development of a security system for homes that uses Computer Vision and Passive Infrared Sensors for detecting Motion. The weighted mean of previous frames was taken, to find substantial changed in motion the thresholded image a limit to the size of contours detected was given and a time limit to see if enough substantial time was passed between the frames. Servo motors were placed in a pan-tilt manner so that the camera mounted on it could be rotated both vertically and horizontally. The system had the capability to be controlled wirelessly via a server which was setup.