Jigar Mehta

Seasoned data analytics professional, Jigar is actively seeking full-time position in data science, analytics and machine learning. I am graduating with M.S. in Analytics in Dec 2016 from J. Mack Robinson College of Business, Georgia State University.

Prior to masters, I have worked for 4 years providing advanced analytics consulting, statistical data analysis and building predictive models for Fortune 500 companies. I have implemented numerous machine learning and text mining algorithms using SAS, R and Python.

My areas of interest include digital marketing, multi-channel attribution, customer analytics, big data analytics, designing and implementing analytical solutions. For my achievements and projects, visit my GitHub.


Find out more

Education

J. Mack Robinson College of Business,
Georgia State University


Master of Science in Analytics
Aug 2015 - Dec 2016
CGPA: 4.0/4.0

Distinct Recommnedation from Dean and Professors
Rank Top 2 in the graduate class


Sardar Patel Institute of Technology


Bachelor of Engineering (B.E.) in Electronics Engineering
Jul 2009 - May 2012
CGPA: 3.9/4.0

Gold medal for securing 1st rank
Tata J.R.D scholarship for academic excellence

Experience

Alliance Data Systems Corporation
(Epsilon Marketing Solutions)

Analytics Intern

May 2016 – Present

- Developing Value, Attrition and Potential segmentation model to achieve incremental ROI for a major automotive client
- Working on logstic regression based response and propensity models to target customers for higher response rate
- Provided optimal resource allocation strategies using marketing mix model, resulted in 12% incremental service revenue

Technology used: SAS, SAS enterprise miner, SQL, Python

Georgia State University

Graduate Research Assistant

Aug 2015 - May 2016

- Topic Modeling and Document Clustering for MIS Quarterly Journal and LDA technique using R and Python
- Temporal analysis to see trend of topics evolving over time using Text Mining and word clouds of keywords
- Moduralized Python code to automatically download and fetch metadata from pdf journals



Altisource Labs

Senior Analyst

Apr 2013 – Jul 2015

- Developed regression models in R that lead to 9% increase in the quarterly collection revenue for a US based loan servicer client with annual revenue of $2.2 billion (includes Model Validation, deployment & testing)
- Achieved 10% reduction in annual lending cost by mortgage default prediction model using regression & survival models in R
- Increased MoM data adoption rate by 3% using decision tree and logistic regression models built in SAS and SPSS for India’s largest telecom client having annual revenue of $33 billion

Awards at this Job:

- Rising Star For Developing and deploying propensity models and exceeding work expectations at client location.
- Young Achiever For Outstanding overall performace in Q4 FY 15

Mu Sigma Inc.

Business Analyst

Jul 2012 – Apr 2013

- Developed GLM Model in SAS to predict and optimize in-store staff needs. Provided framework to analyze store operations, traffic patterns and inventory management for Fortune 33 brand with an annual revenue of $75Bn
- Presented dashboard in Tableau and insights using strategic consulting - Quarterly Revenue Forecasting and Growth numbers, National, Regional & State level comparison of business KPIs, Performance benchmarking

Awards at this Job:

- Spot Award For Developing an automation tool for payroll processing that reduced the weekly man hours by 80%

Recent Projects

Credit Scoring for Mortgage Applicants using SMOTE and Gradient Boosting

  • Summary : To classify the loan applicants as defaulters or not based on the ensemble of machine learning algorithms like probabilistic model, boosting algorithm, classification tree methods. Used Synthetic Minority Over-Sampling Technique (SMOTE) to handle imbalanced classification. Parameter tuning using Grid search was used over greedy algorithm for Gradient boosted trees algorithm.
  • Tools and Technologies used : Python , R
  • Libraries used : scikit learn, SMOTE-unbalanced
  • Click on this link for GitHub

Sentiment and Opinion Mining of Amazon and eBay product reviews using Tweets

  • Summary : Mining live twitter feed data about two companies amazon and ebay to analyze the sentiment of each tweet using Naive Bayes Model and aggregating at US states level. Displaying results on map using geo-locations. The results include identification of top keywords used, top users and top trending topics. Novel technique of model output as feedback was used to self tune the model and increase the accuracy.
  • Tools and Technologies used : Python, Tableau
  • Libraries used : NLTK, Python Streaming Twitter API
  • Click on this link for GitHub

Predicting Impact of New Messaging feature on Booking Experience for Airbnb

  • Summary : Determine the impact of length of messages sent, as one of the differentiating factor, on guest booking experience. Determing partial effect of message length of place bookings negating the effect of other parameters. A/B testing and control groups was used to determine the results.
  • Tools and Technologies used : SAS, R
  • Click on this link for GitHub

Topic Modeleling and Movie Genre Prediction using Natural Language Processing

  • Summary :. Movie reviews were stored in MongoDB in json format. Textual analysis was performed on the synopsis of the reviews to find few search results. Supervised machine learning algorithms were deployed to predict genre classification coupled with topic modeling output. The LDA Topic Model algorithm was tuned using multiple variations of TF-IDF options. Automated model comparison and scoring method was implemented.
  • Tools and Technologies used : SAS, SAS Enterprise Miner, MongoDB
  • Click on this link for GitHub

Skills


Professional Certifications

>> MACHINE LEARNING_UNIVERSITY OF WASHINGTON

>> MACHINE LEARNING_STANFORD UNIVERISTY

>> SAS CERTIFIED STATISTICAL BUSINESS ANALYST

>> SAS CERTIFIED ADVANCED PROGRAMMER

>> SAS CERTIFIED BASE PROGRAMMER

>> BUSINESS ANALYTICS - EDUPRISTINE

Skills

Analytical Tools and Softwares

  • SAS

  • R

  • Python

  • SQL/PL SQL

  • Tableau

  • SPSS Statistics / Modeler

  • Visual Basic / VBA

  • Hadoop Ecosystem, MapReduce

  • Apache Spark

  • SQL Server DB Services (Integration, Reporting and Analysis services)


Machine Learning Algorithms

  • Linear and Logistic Regression

  • LASSO and Ridge Regression - Regularization


  • Recommender Systems


  • Ensembles methods - ADABoost, Gradient Boosting, XGBoost, Random Forest




  • Nearest Neighbours


  • Topic Modeling, Text Mining


  • SVM - Linear and Gaussian Kernal


  • K-means and Hierarchical Clustering


Analysis Experience

  • Sentiment Detection

  • Anomaly Detection

  • Dimensionality Reduction - PCA

  • A/B Testing

  • Response, Propensity Models

  • Lifetime Value, Media Mix and Loyalty Analysis


  • Cross-Sell & Up-Sell Models

  • Telecomm, Retail analytics

  • Web and Digital Analytics

  • Data Audits, Data Governance


Contact Me

Address 2399 Parkland Drive, Atlanta, GA 30324, USA

Phone +1 678-779-8636

Email jigarmehta2@gmail.com

Alt Email jmehta3@student.gsu.edu