Booking open Available Pune Call Girls Ambegaon Khurd 6297143586 Call Hot In...
Vivek Adithya Mohankumar - Resume
1. VIVEK ADITHYA MOHANKUMAR
402, Kerby St, Timberbrook Apt 216, Arlington, TX 76013
682-365-0516 | vivekadithya.m@gmail.com | vivekadithya | vivekadithya
EDUCATION
University of Texas at Arlington Arlington, Texas
MASTER OF SCIENCE IN INFORMATION SYSTEMS (GPA : 3.66/4.0) Expected May 2017
Anna University Tamilnadu, India
BACHELOR OF INFORMATION TECHNOLOGY July 2008 - May 2012
WORK EXPERIENCE
Information Developer
SAP LABS INDIA PRIVATE LIMITED December 2012 - July 2015
• Worked with product managers and analysts, on product requirement analysis and specification documentation.
• Analyzedtargetusergroupsandauthoreduserguides, systemconfigurationguides, userinterfacetextsandothertrainingdocuments.
Delivered documentations to support features of 14 countries within 3 months.
• Documented and presented products at technology conferences to business partners and customers.
Graduate Research and Teaching Assistant
DEPARTMENT OF MANAGEMENT, UNIVERSITY OF TEXAS AT ARLINGTON June 2016 - August 2016
• Assisted Dr. Wendy Casper in a research study on cultural adaptability of foreign students in the United States. I was responsible for
the experiment design, data collection and processing.
COMPUTER PROFICIENCY
Big Data and Cloud Computing : Python, SQL, Hadoop, PySpark, SparkSQL, SQOOP, Hive, Impala
Text Analysis and Machine Learning : Scikit-learn, Gensim, Pandas, TextBlob, NLTK
Data Visualization : Tableau, D3.js, ggplot, Matplotlib, SAP Lumira
Statistical Data Analysis : SAS, Eviews, STATA, R, Microsoft Excel
Other Tools and IDEs : SAP Netweaver, MS Office, Microsoft Excel - Solver, Pycharm, iPython Notebooks
PROJECTS
• PySpark: Music Artist Recommendation on Yahoo! Music Data:
A PySpark application based on ALS algorithm was developed to recommend a user music artists. The data cleaning process involved
scaling down the ratings to a range of 0 to 5 and removing the outliers that skew the modeling. The data was split in the ratio 0.6:0.2:0.2
for training, validation and testing, with the validation and testing sets containing only the user and artist ids. An ALS based model was
built and the performance was evaluated on RMSE metric. A new user was hardcoded to the original dataset with random ratings to
eight artists. A model was trained on this updated dataset and scores for other artist’s were predicted. Finally, 5 artists with the best
predicted ratings were suggested to the user.
• MapReduce Application to Analyze On-time Performance of US Domestic Flights:
A Hadoop Map Reduce application to report maximum departure delay for each originating airport, average arrival delay by flight
number and minimum arrival delay for all origin-destination airport combinations. Individual mappers and reducers were developed
for each task. The intermediate data from the mapper was sorted and sent as input to the reducer.
• Yahoo! Answers Best Answer Validation:
The project validated the accuracy of best answer selection procedure. Used NLTK to preprocess the corpus and deployed machine
learning algorithms using Sklearn module to model the validation engine. Some of the analysis were estimation of the answer length
(best answers in most cases are written by combining the best parts of the answers posted earlier), estimation of cosine-similarity,
estimation of jaccard similarity and topic modeling(using Gensim) to cluster similar answers.
• Twitter Text Analysis - US Presidential Candidates:
The project involved Sentiment Analysis of the tweet corpus to understand the sentimental influence of each presidential candidate.
We studied the tweets based on the states, to understand the candidates influence with respect to the states. We also created a word
cloud to display the top 50 words that appeared in the overall tweets to give a perspective of how people react to politics on social
media.
LIKES
• Medium, Ycombinator News, Quora, Kickstarter Projects, Binge-watch Youtube, a whiteboard and some crazy ideas