Data Scientist

Rohit Dandona

Bloomington, IN, USA aczchf@r.postjobfree.com +1-812-***-**** https://www.linkedin.com/in/rohitdandona https://github.com/rohitdandona

Education

MS, Data Science, Indiana University, Bloomington, IN, USA (Expected) May 2017 BE, Information Science, Visvesvaraya Technological University, Karnataka, India June 2011 Professional Experience (4 years)

Indiana University, Teaching Assistant, USA Jan 2017 – Present

• Conduct office hours, grade assignments, exams for the SQL/NOSQL course. Develop an online Hadoop course module. Exusia, Inc., Senior Analyst, India/South Africa Apr 2015 – Dec 2015

• Designed, installed and administered a MapR Hadoop infrastructure for a client in the banking sector. Part of the ETL model architecture development of the client’s data processing migration from Data Stage/Mainframe to a Hadoop/Abinitio space Infoavent Consulting Ltd., Senior Software Developer, India Nov 2014 – Mar 2015

• Designed and created a Hadoop environment for media web-log data processing for real-time/non-real-time reporting Syntel, Inc., Software Developer, India Mar 2012 – Oct 2014

• Worked extensively with Hadoop, Java MapReduce and Hive Query Language, on three data warehousing projects while building an Enterprise Customer Record platform. Gained comprehensive hands-on experience in Hadoop based software development for complex business scenarios in the retail space, and in the over ETL lifecycle.

• Awarded Syntel “SMART” value award twice for outstanding performance Data Science Projects

Prediction and Analysis of Crime in San Francisco – Kaggle [Scikit-Learn, Python, R, Tableau] Spring 2016

• Analyzed and visualized the spatial/temporal relationship of crime to predict the category of crime in a particular location. Implemented a Decision Tree, a Random Forest and a Support Vector Machine classifier. Classification Algorithms without using libraries [Python] Fall 2016

• Implemented Decision Tree, Naïve Bayes classifier, Adaboost, Neural Network, K- nearest neighbor algorithm on various UCI datasets for use cases like movie recommendation, spam detection, image orientation classification etc. Regression Algorithms without using libraries [Python] Fall 2016

• Deployed variants of linear regression featuring Batch/Stochastic gradient descent, l1 and l2 regularization, feature selection, etc. on a UCI dataset of CT slices

Yelp Dataset Challenge [MongoDB, Python, Java] Fall 2016

• Built a model to predict business category with “Review” and “Tip” data of the Yelp dataset. Employed the Bag of Words approach using TF-IDF scores to extract features and classify using Naïve Bayes. POS Tagging using Hidden Markov Models [NLP, Python] Fall 2016

• Performed Parts of Speech Tagging using Hidden Markov Model (HMM) and higher order HMM using Viterbi Algorithm and Dynamic Programming.

Image Sentiment Analysis on Facial Expressions [Scikit-Learn, Python, R, Tableau] Summer 2016

• Developed an effective neural network architecture featuring several fully connected layers to analyze and predict image sentiment labels. Implemented a SVM classifier using SIFT descriptors for comparison as well. Routing system [Python] Fall 2016

• Designed and developed a routing system using various search techniques (BFS, DFS, IDS, A*) wherein the user can opt to be guided via various routing options such as least turns, least distance, least time or one which is the most scenic. Technical Skills

Analytical/Inferential Skills: Predictive Modelling (Classification and Regression), Exploratory Data Analysis, Feature Engineering, Hypothesis Testing, ANOVA, Text Mining NLP: POS Tagging, Word2Vec Modelling, Topic Modelling, Sentiment Analysis Languages/Tools: Python (Scikit Learn, Pandas, Numpy, Matplotlib, NLTK), R, Java, HQL, Shell, C/C++, Tableau Big Data/Database Skills: Hadoop, MapReduce, Hive, Pig, Spark, Flume, Kafka, SQL, MongoDB, HBase, Neo4j

Contact this candidate