Open

I am Aditya Anand Kavale. I have completed my schooling in Bangalore and currently in second year of college studying Computer Science Engineering in RV University Bangalore, India. I have always had a huge interest in Astronomy therefore making sure to do a project involving or related to astronomy using the HPCC Systems. I have been coding since High school and have gained an interest in it and is one of the reasons I chose Computer Science Engineering. Coding in HPCC Systems using ECL was something new and quite an experience which I really loved. My hobbies include watching sports, reading and jogging. I also really love reading information or topics related to astronomy. I am really looking forward to use HPCC Systems in the future and gain more knowledge on it and ECL language.

Poster Abstract

Asteroid detection and tracking are essential for safeguarding Earth from potential collisions with hazardous asteroids, which could have catastrophic consequences. Large asteroids have the potential to cause massive explosions, tsunamis, and significant climate changes, as evidenced by the asteroid impact that contributed to the extinction of the dinosaurs. Even smaller asteroids can lead to significant damage and loss of life if they strike populated areas. Early detection and tracking enable the development of evacuation plans and the implementation of mitigation strategies, such as deflecting the asteroid's course.

HPCC Systems is a powerful platform for handling large-scale data processing tasks, complex data transformation and integration needs, real-time data analytics, and data mining applications. Its parallel processing capabilities allow for efficient management of massive datasets, while its comprehensive ETL engine supports intricate data extraction, transformation and loading from diverse sources. It helps in storing large chunks of important data related to the asteroid which includes eccentricity, asteroid diameter etc.

Furthermore, HPCC Systems also features machine learning libraries that can scale up to help, build and deploy predictive models and applications for data mining. As an open-source solution, it provides a cost-effective, flexible and secure environment for transforming raw data into actionable insights.

The primary objective is to improve the accuracy and efficiency of identifying and tracking potentially hazardous asteroids. Machine learning algorithms are applied to the refined data. Using the GLM bundle, HPCC Systems precisely identified hazardous asteroid patterns through classification techniques. This allowed for accurate prediction and analysis of potential asteroid threats. HPCC Systems helps in cleaning the dataset by removal of unwanted data and providing with mean of various asteroids diameters in different measurements. It also accommodates machine learning bundles as well as offers integration with python which provides TensorFlow library for deploying machine learning models which gave an accuracy of 84.328% for our target column named 'Hazardous' using Logistic Regression, which is a generalized linear model (GLM) commonly used for binary classification tasks. This integration demonstrates a significant improvement in the speed and precision of asteroid detection and tracking. The results indicate that this combined approach provides more accurate predictions and faster processing times compared to traditional methods. The discussion highlights the system's scalability, flexibility, and potential for real-time application in planetary protection strategies.

Future work will focus on enhancing the machine learning models with more extensive training datasets and incorporating additional data sources to improve prediction accuracy.

Presentation

In this Video Recording, Aditya provides a tour and explanation of his poster content.

Asteroid Detection and Tracking:

Click on the poster for a larger image.

Open