EKTA RAJIO AKADKAR

+1-585-***-**** adjlzx@r.postjobfree.com LinkedIn - /ektaakadkar/ GitHub - /ektaakadkar Rochester (NY), USA EDUCATION

Rochester Institute of Technology Expected: Dec 2021 Master of Science - Computer Science GPA: 3.63/4

Relevant Coursework: Data Warehousing, Big Data Analytics, Advanced Object - Oriented Programming, Foundations of Algorithms, Foundations of Artificial Intelligence, Foundations of Computer Vision University of Pune Aug 2015 - Jun 2019

Bachelor of Engineering - Electronics and Telecommunications GPA: 8.24/10 Relevant Coursework: Machine Learning, Robotics, Data Structures and Algorithms, Fundamentals of Programming Language TECHNICAL SKILLS

Languages: Python, R, Java, NodeJS, JavaScript, HTML, CSS Databases and Data - frames: MySQL, MongoDB, Pandas, NumPy Frameworks and Tools: Pentaho (Kettle), SAS, MySQL Workbench, R Studio, Tableau, Gradle, Google Collaborator, MATLAB (Simulink) Machine learning models: Generative Adversarial Networks, Long Short - Term Memory, Multi - label Classifier (Neural Network), Decision Tree, Random Forest, Naive Bayes, Support Vector Machine, Linear Regression, Logistic Regression, KMeans, KNN, Apriori, Principal Component Analysis

Libraries and Packages: ggplot2, twitterR, dplyr, tidytext, tm, udpipe, sentimentr, BeautifulSoup, matplotlib EXPERIENCE

Graduate Research Assistant: [ PyTorch, Python, Google Collaborator ] Jul 2020 – Present

• Worked on improving the accessibility of voice assistants relying on aural interfaces for deaf (DHH) users by implementing a machine learning model for monotone to normal speech conversion with Generative Adversarial Networks and Griffin - Lim algorithm on self - collected raw dataset in PyTorch.

• Comparing style transfer methods implemented on different researched audio signal processing techniques.

• Classified audio signals consisting of 4 human voices speaking numbers with an accuracy of 83 percent by implementing CNN on a customized data set.

Aker Solutions: Project Intern Aug 2018 – Jun 2019

• Implemented the code to navigate the Remotely Operated Vehicle (ROV) underwater by collecting data from sensors and actuators and processed it on microcontrollers. The entire electronic system was also designed and implemented. Publication : Design for Navigation of Underwater Remotely Operated Vehicle, IRJET 30 Sept 2019 See Publication PROJECTS

Data mart implementation using ETL process: [ MySQL, Pentaho (Kettle), SAS ] See Project Oct 2020 – Nov 2020

• Increased the profit margins of inter - related companies by providing a self - designed sales data mart in MySQL using dimensional modelling techniques by performing ETL operations in Pentaho (Kettle) on 3 OLTP databases.

• Implemented the data mart consisting of 6 dimensional and 1 fact table using STAR schema and constructed the E - R model.

• Aggregated data marts were developed for certain use cases using shrunken, collapsed and lost dimensions.

• Cleaned, reformatted and maintained uniformity in the data using SAS. Sentiment analysis of ‘vegan’ tweets: [ R, Tableau ] May 2020 – Jul 2020

• Comprehended perceptions of people practicing veganism and concluded that positive emotions influence health conscious people by 15 percent more by performing sentiment analysis on 0.5 million of tweets.

• Cleaned and quantified the polarity of tweets based on sentiments. Further, classified the tweets using Naive Bayes and Random Forest algorithms into topics obtained by Hierarchical clustering and Latent Dirichlet Allocation.

• Performed exploratory data analysis (EDA) to gain more insights on rise in veganism, density plot of locations and probability of retweets.

ChatBox client for browser: [ NodeJS, Express, Mongoose (MongoDB), jQuery ] May 2020 – Jun 2020

• Developed a ChatBox app hosted on a website for multiple users featuring ‘http’ requests and created web sockets for automatic message notifications. Connected MongoDB Atlas at the backend to store and retrieve the message data. Computation on IMDb movie dataset: [ Java, JDBC, MySQL, MongoDB ] Jan 2020 – Feb 2020

• Transferred the data using JDBC driver into a top - down approach of denormalized relational model.

• Decreased data redundancy by computing functional dependencies, candidate keys and canonical covers on the IMDb movie dataset consisting of 50 million rows.

• Transferred data from MySQL to MongoDB and retrieved data by applying advanced queries on both the database formats Language classifier: [ Python, Pandas ] See Project Jan 2020 – Feb 2020

• Deduced 10 differentiating factors after analyzing the language conformation and classified short sentences into English and Dutch languages with an accuracy of 92 percent by building Adaboost and Decision Tree classifier models trained on 10000 lines of dataset.

Path finder using A* algorithm: [ Python, OpenCV ] Jan 2019 – Feb 2020

• Constructed the fastest path finder for a park map taken as an input image (385 x 500 pixels) consisting of 10 different terrain by implementing A* algorithm. The condition of terrains is considered as per the season.

• Formulated the cost function of the algorithm by deciding the heuristic function based on the speed of a person on the terrain and the time required to cover the distance between two points. Battleship game using TCP: [ Java ] See Project Oct 2019 – Nov 2019

• Developed a battleship game for 2 players connected to different servers communicating bi - directionally with each other to convey the result of each move by implementing TCP connection in Java. ACTIVITIES AND ACHIEVEMENTS

MIT Tech Team ( National Awards ) : Technical Head Aug 2016 – Apr 2018 ABU ROBOCON 2017: 2nd place out of 110 universities ABU ROBOCON 2018: 2nd place out of 103 universities SAVe NIOT 2019: 3rd place out of 169 universities

Contact this candidate