Deep Learning

Hurricane path prediction from HURDAT dataset

Every year hurricanes cause damages in the USA and all around the world at a huge scale. Hence, any prior knowledge of a hurricanes path will facilitate in predicting the region where it might be headed next and help in taking precautionary mea- sures beforehand. We can exploit the dataset named HURDAT database, provided by the National Oceanic and Atmospheric Administration (NOAA) to feed it into different supervised learning models. Subsequently, a comparison of the accuracy of the result predicted is discussed.

Vehicle Detection and Tracking

Detecting car in a video sequence and tracking the car in the video sequence. A classifier is trained for detecting car in an image and the neural network using for tracking the car in the sequence using sliding window technique