Hi! I am Sonal

I am a Data Scientist with 4 years of experience in solving real business problems using Machine Learning and Predictive Analytics

Tell me more!

Passionate, skilled and creative problem solver

I enjoy building data driven solutions to solve complex problems. I worked in one of the leading Data Science Companies, Dunnhumby for four years. Throughout my tenure, I was involved in designing Analytical Customer Relation Management mechanisms and solutions for Loyalty Programs for various organizations and helped retailers make smarter decisions through data-driven insights.

Core Competencies

Business

Understanding business and customizing solution for the client while communicating clearly with the cross-functional team, stakeholders, and users

Data

Experienced in mining structured/unstructured data and extracting meaningful insights

Machine Learning Analytics

Supervised Learning (Linear Regression, Logistic Regression, KNN, SVM, Naïve Bayes, Decision Trees, Random Forest, Artificial Neural Network), Unsupervised Learning(K-Means, Hierarchical Clustering, Dimensionality Reduction Techniques), Control Selection, Hypothesis Testing, Statistics, Sentiment Analysis, Collaborative Filtering, Biased Sampling

Code

Python, R, PySpark, SAS, SQL, MATLAB, C

Sentiment Analysis on Yelp Data
  • Toolkit: Python, Logistic Regression Model, Random Forest, SVM, n-gram sequences, vectorizers
  • Concept: Predicted the rating on Yelp by performing Sentiment Analysis on review
  • Data: Business and Reviews dataset obtained from official website of Yelp as part of Yelp Data Challenge
Classification and Supervised Learning on Financial Data
  • Toolkit: R, Decision Tree, Random Forest, dplyr, ggplot2
  • Concept: Analysing the Financial data and predicting whether the S&P 500 Index would have a gain for two days
  • Data: Stock Prices with Open, High, Low and Close values for a day with a binary target variable
Web Scraping
  • Toolkit: Python, Beautiful Soup
  • Concept: Scraped the Indeed website to gather information around skills needed for a Data Scientist roles in top companies as well as Reviews of top companies
  • Data: Scraped from the website
Predicting West Nile Virus in Chicago
  • Toolkit: Python, Logistic Regression, XG Boost, Random Forest, SMOTE, Feature Engineering
  • Concept: Obtained the likelihood of West Nile Virus being present given the spray time, location, weather and mosquito species
  • Data: Created a master dataset collating information obtained from Weather,Spray and Trap across multiple years
Won First Prize in the Hackathon organized by Orange County's R Users group
  • Toolkit: R, Logistic Regression, SMOTE
  • Concept: Predict whether the client will subscribe (yes/no) to a term deposit. Key Process Owners can use the output of this prediction to improve the stragegy for the next market campaign
  • Data: The data is about direct marketing campaigns (phone calls) of a Portuguese banking institution from May 2008 to November 2010. The clients were contacted more than once in order to access if the product (bank term deposit) would be (‘yes’) or not (‘no’) subscribed.
Clustering and Unsupervised Learning on Airbnb LA data
  • Toolkit: R, K-Means, Multivariate Linear Regression, dplyr, ade4, ggplot2
  • Concept: Obtained the likelihood of West Nile Virus being present given the spray time, location, weather and mosquito species
  • Data: Created clusters of similar Airbnb rooms in Los Angeles, California and Built a multivariate linear regression model to predict the price of a room in each cluster
Analysis of Stock Prices and Crunchbase database
  • Toolkit: Python, SQL, Seaborn, Matplotlib, Cufflinks, Plotly, Iplot
  • Concept: Analysis of Stock Prices for major banks and Crunchbase database
  • Data: Stock data collected using DataReader package in Pandas
Gun Violence incidents in the United States
  • Toolkit: Tableau
  • Concept: Analysis of Gun Violence Incidents in America in 2017
  • Data: Publicly available data containing details around every gun related incident in the United States
Price Prediction
  • Toolkit: R, Multivariate Linear Regression
  • Concept: Predicting the price of a house in King County, Washington
  • Data: Information around the houses sold in King County between May 2014 and May 2015

Let's Get In Touch!

Feel free to contact me if you would like to learn more about any of my projects or simply chat about data science or why we all hate pie charts.