Authors: Jayant Mehra
Although the current detection techniques have been able to calculate the minimum distance to which a Near Earth Object (NEO) can approach Earth for thousands of NEOs, there are millions of yet undiscovered NEOs which could pose a threat to Planet Earth. An NEO is considered highly dangerous if the minimum distance between it and the centre of the Earth is less than 0.03 AU. However, only a handful NEOs have been detected prior to entering this danger zone. The immense task of asteroid hunting by conventional techniques is further complicated by a high number of false positives and false negatives. In this report, machine learning algorithms are written to predict the minimum distance upto which an NEO can approach the planet and classify NEOs as whether they are in the danger zone or no based on their physical characteristics. In section 4 of the study, an Artificial Neural Network based on the backpropagation algorithm and a Logistic Classification based on Unconstrained Minimisation using the fminunc function are employed to classify NEOs with an accuracy of 92% and 90% respectively. In section 5 of the report, the Levenberg - Marquardt Algorithm based on an Artificial Neural Network is employed to calculate the minimum distance with a regression R value of 0.79 (Value of 1 being the maximum). All the algorithmic systems developed have low false positive and false negative rates
Comments: 15 Pages, 6 Figures, 5 Tables
[v1] 2016-12-30 06:29:12
Unique-IP document downloads: 13 times
Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.