Kalpesh Adhatrao
1226 W Adams Blvd. Apt. 3C, Los Angles, CA • 213-322-9429 • kalpesh.adhatrao@gmail.com
Hi there! This is Kalpesh Adhatrao. I am a Graduate Student pursuing MS in Computer Science(Data Science) at the University of Southern California. Before joining USC, I worked as a Senior Software Engineer for a US-Based client named Health Fidelity with the focus on delievering NLP and Analytics solutions to the Health Care system.
Experience
University of Southern California
Graduate Student Researcher • Aug, 2018 — PresentWorking on a Data-Driven Journalism based project under Prof. Yao-Yi Chiang to uncover interesting patterns from Los Angeles crime data.
University of Southern California (Graduate School)
Student Assistant Programmer • Jan, 2018 — Nov, 2018Led the development of an administrative dashboard to manage, model and analyze graduate students’ data.
WebShar India Pvt. Ltd.
Senior Software Engineer • June, 2013 — Nov, 2017- Achieved a performance gain of 50% by rewriting the existing Rule Engine module to generate highly relevant clinical coding suggestions for clinical phrases in the medical notes.
- Improved the precision and recall metrics of the NLP engine to the range of 90-95% by developing a Java-based pre-processing engine to identify sentences, sections, list items etc. from medical notes.
- Collaborated in the development of a platform to monitor and process a large number of medical notes with the help of CDH (Cloudera Distribution Including Apache Hadoop).
- Developed a search service using Apache Lucene to get clinical coding suggestions for the given clinical phrase.
- Built a Topic Modeling system to derive categories from medical notes using the implementation of Latent Dirichlet Allocation algorithm from scikit-learn
- Implemented and evaluated classification models using scikit-learn to identify risk level associated with patients based on their medical history.
- Engineered a medical notes encryption pipeline in Java using AWS Encryption SDK.
- Spearheaded the development of an administrative dashboard using Spring MVC & AngularJS to manage access levels of all the applications in the platform.
Education
University of Southern California
Master of Science in Computer Science (Data Science) • Expected - Dec 2019
Coursework:
- Analysis of Algorithms
- Foundations of Artificial Intelligence
- Foundations and Applications of Data Mining
- Applied Natural Language Processing
University of Mumbai
Bachelor of Engineering in Computer Engineering • Aug 2009 — May 2013
Relevant Coursework:
- Discrete Structures & Graph Theory
- Data Structures & Files
- Database Management System
- Analysis of Algorithms & Design
- Advance Database Management System
- Theory of Computer Science
- Web Engineering
- Data Warehouse & Mining
- Robotics & AI
- Distributed Computing
- Software Architecture
Projects
- Recommender System: Collaborative-filtering based movie recommendation system built using the MovieLens dataset.
- Performance Prediction: Web application developed in CodeIgniter that uses models trained with ID3 and C4.5 algorithms to predict the performance of undergraduate students based on their past academic & personal details.
- Placement Portal: A PHP-based web application to collect and manage the academic and personal details of students and generate campus recruitment specific statistics.
- BuyIt: Web application built using React, Express & MongoDB to track savings for the items in the wishlist.
Skills
Languages
Java, Python, Scala, C#, SQL, PHP, JavaScript, HTML & CSS
Technologies
Spring Boot, CodeIgniter, ASP.NET, jQuery, React, Kendo UI, AngularJs, MongoDB, SQL Server, Docker, AWS, Apache Spark, Apache Lucene, IPython, Pandas, NumPy, Matplotlib, Scikit-learn
Workshops & Talks
-
Delivered a lecture on topic Introduction to Machine Learning in Python at DBIT, Mumbai in April 2017
Link: https://github.com/kelsadita/Machine-Learning-Talk -
Conducted a workshop on topic Data Science and Industry for the short-term training program on Data Science and Big Data Analytics at FCRIT, Vashi in December 2016
Link: https://github.com/kelsadita/DataScienceSTTP
Publications
-
Predicting Students’ Performance Using ID3 and C4.5 Classification Algorithms Abstract: An educational institution needs to have an approximate prior knowledge of enrolled students to predict their performance in future academics. This helps them to identify promising students and also provides them an opportunity to pay attention to and improve those who would probably get lower grades. As a solution, we have developed a system which can predict the performance of students from their previous performances using concepts of data mining techniques under Classification. We have analyzed the data set containing information about students, such as gender, marks scored in the board examinations of classes X and XII, marks and rank in entrance examinations and results in first year of the previous batch of students. By applying the ID3 (Iterative Dichotomiser 3) and C4.5 classification algorithms on this data, we have predicted the general and individual performance of freshly admitted students in future examinations.
-
A Secure Method for Signing in Using Quick Response Codes With Mobile Authentication Abstract: The emerging threats to user privacy over the internet are increasing at an alarming rate. Signingin from an unreliable terminal into a web account may result in compromising private details of a user such as username and password, by means of keylogger software. Such software are capable of recording keystrokes secretly, via covert channels without the knowledge of the user. In this paper we propose a secure method for signing in using Quick Response (QR) codes with mobile authentication. Through this method, the user can securely sign-in into a web account by authenticating the user session on an unreliable terminal browser, using a mobile device.
-
Parallelizing Network Flow Using ϵ-relaxation Algorithm Abstract: Maximum flow problem is a fundamental network flow problem which involves finding a feasible and optimal flow in network consisting of single source and single sink. It has wide ranges of research and commercial applications. Many sequential algorithms such as Ford-Fulkerson algorithm, Edmonds-Karp algorithm, Goldberg’s “generic” maximum flow algorithm, ϵ-relaxation algorithm have been designed to solve this problem. In this paper, we present parallel implementation of ϵ-relaxation algorithm which has been carried out over a cluster consisting of 1 master and 4 slaves nodes in order to study and improve the efficiency of solving maximum flow problem on multi-computer architecture.
Social Links
- Github: https://github.com/kelsadita/
- Twitter: https://twitter.com/kelsadita
- LinkedIn: https://www.linkedin.com/in/kalpeshadhatrao/