Deep Learning and Anomaly Detection in Mars Rover Data

About this technology

Overview

The included software is for the data processing and machine learning analysis used inthe project Deep Learning and Anomaly Detection in Mars Rover Data Transmission. The filescontained the signal_processing directory contain Python scripts to organize and clean a datasetbefore passing it into the various machine learning and deep learning models. Themachine_learning directory contains an implementation of an Adversarial Autoencoder, a deeplearning model used in reconstructing data, and a file for a One-Class SVM, which is used indetecting anomalies in the dataset. It also contains a deep neural network implementation, whichis used in classifying the success of different Mars downlink successes.