google-site-verification: googlefaee0f3fb2c07645.html Aishwarya Rajasekaran

I'm a first year MS in Computer Science student working towards Data Science Concentration at the University of Massachusetts, Amherst. My research interests are broadly inclined towards building Trustworthy AI from both technical and ethical perspectives. I believe that we need to look beyond numerical performance metrics to evaluate AI systems. Social and ethical parameters should also play a significant role during decision making.

I completed my B.Tech in Computer Engineering with Distinction from Indian Institute of Information Technology, Design and Manufacturing Kancheepuram in 2019. I was fortunate to spend a semester and complete my Bachelor Thesis at IIT-Madras under the guidance of Prof. Balaraman Ravindran.

Cooking, playing violin, travelling, mentoring, doodling, watching cartoons and playing basketball keeps me occupied when I am bored.

Professional Experience

May 2018 - Oct 2018

Research Intern

Guide : Dr Balaraman Ravindran

I worked on "Development of a Deep Learning Algorithm for Detection of Retinopathy in Retinal Fundus Photographs". The main categories are Diabetic and Hypertensive retinopathy. Interpreting these fundus photographs requires specialized training, and in many regions of the world there aren’t enough qualified graders to screen everyone who is at risk. This is where AI can help doctors identify patients in need, particularly among underserved populations.

May 2017 - July 2017

Developer Intern

Guide : Dr Amaresh Chakrabarti

Worked as a Full Stack Web Developer and designed the Database for InDeaTe, a Computer-Based Platform with a Systematic Design Template and a Database of Methods and Tools which aims at improving sustainability considerations in design.

May 2016 - June 2016

Summer Intern

Implemented encryption and decryption of text files using DES algorithm in JAVA

Projects

EXPLAINABLE NN FOR IMAGE CLASSIFICATION WITH TRUST METRIC


Accountability and explanation of the decision is crucial in safety critical applications. The proposed method is simple and provides a post-hoc explanation for model’s decision with a confidence score.

MINI SEARCH ENGINE


Built a search engine in Apache Spark to create inverted index of urls. The input files were stored in HDFS. RocksDB was used for storing the inverted index and serve queries. Finally, Flask server was set up to serve the queries.

IMAGE CLASSIFICATION SERVER






Implemented an image classification server using pretrained Densenet-121 model. The server was set up using Flask and a Docker container was built to use the model for inference.

FRS BASED PORTABLE ATTENDENCE SYSTEM


A portable attendance system to manage the student attendance at IIITDM. It is based on feature matching algorithm that runs on raspberry pi.A website was also developed to view the generated reports.

HEART RISK PREDICTION BASED ON DATA MINING TECHNIQUES


Used data mining techniques like Linear Regression and Decision Tree to recognize and correlate various cardiac parameters with their corresponding medical conditions.

FILTER FOR INDUSTRIAL GASEUOUS EFFLUENTS


To reduce the amount of gaseous effluents released by industries, an efficient dynamic filter was designed with inbuilt air quality monitoring system.

"Design was selected as one of the best 3 out of 80 projects”

Professional Skills
Programming and Scripting Languages:

C, C++, Python, MATLAB, HTML, SQL

Frameworks and Tools:
Pytorch, Django, Apach Spark
Leadership Skills
I started the Mentor-Mentee Program at IIITDM and lead a team of 30 Mentors and 250 Mentees
Project Management
Lead the Team Nicotop at CIAP-AICTE Startup Mentorship Program -2018 and were in top 15 among 780 teams across India
Achievements