Isha Arora Email & Phone Number

Boston, Massachusetts, United States

Boston, Massachusetts, United States

● Reviewing and enhancing state-of-the-art deep learning models for grading prostate cancer using whole slide images. ● Monitoring of performance shows an agreement of 0.66 (quadratically weighted kappa) and 0.81 (weighted accuracy).● Engineered a use-case to extend to Breast Cancer using 9109 WSIs (BreakHis) and RNA-sequencing data (TCGA) for multi-modality.● Attained a 0.95 weighted accuracy on the binary classification of the BreakHis dataset using a CNN-LSTM model.

Boston, Massachusetts, United States

Boston, Massachusetts, United States

Teaching Assistant for DS 3000: Foundations of Data Science

Boston, Massachusetts, United States

● Analyze and research PTSD reporting metrics using statistical modeling and computational methods assessed on data sets retrieved from patients of the hospital.● Correlation between the different reporting metrics was found to be very strong at 0.82. AUC of 0.93 showed excellent diagnostic performance.● Researched an ML based approach using Electronic Medical Records to identify the risk factors for Childbirth - PTSD.

Boston, Massachusetts, United States

Teaching Assistant for DS 2000: Programming with Data

Boston, Massachusetts, United States

Teaching Assistant for DS 2000: Programming with Data

Hyderabad, Telangana, India

● Coordinated with Wolters Kluwer USA to create a system hosting all the laws and regulations in banking and insurance. ● Accelerated the lookup time for laws and the system aimed to help improve client efficiency by at least 60% in 5 months.● Developed and cleaned a database using PostgreSQL for these laws and regulations for all 50 states. ● Refined the data sent in by the client using OpenRefine and assembled it according to the US states and the relevant books.

Thane, Maharashtra, India

● Conceptualized a project independently on spam detection on a set of 4480 reviews for a banking application.● Implemented Naive Bayes, Decision Tree, and the Apriori algorithm for spam detection. Achieved an accuracy of 61%.● Executed the VADER algorithm for sentiment analysis on the reviews to assess the application. At least 50% of the reviews marked as non-spam were positive.

Thane

» Involved in creating a system for logging the tickets raised by the clients and the employees that were sent by e-mail.» Created a database of tokens divided under each department for each project, using the existing tickets. » The existing ticket data was used to analyze the kind of tickets that were previously logged under each department and project. This analysis was then compared with every new ticket that came in by e-mail. This new ticket was assessed using Naïve Bayes and Cosine similarity to decide under which department and project the new ticket should be logged.

Mumbai Area, India

» Wrote SQL queries and retrieved data as required by the company, as the company’s software was being updated to be developed in a new language. Was given daily tasks during the course of this internship.» Cleaned and fixed quality-based issues in the C# code for the health management system developed by the company using SonarLint.

Vellore Area, India

A part of the Management and the Editorial teams at GDG VIT Vellore. As a part of GDG VIT, taken part in the Google Online Marketing Challenge (GOMC ) 2017 representing Notebit.in

Master Of Science - Ms, Data Science, 3.83

Completed: Introduction to Data Management and Processing (DS5110) Supervised Machine Learning (DS5220) Unsupervised Machine Learning.

Bachelor Of Technology, Computer Science, 8.67/10

Google Developers Group (GDG) VIT Vellore Chapter Make A Difference (MAD) Vellore Chapter

Higher Secondary School Education

Secondary School Education, Middle School Education