Data Scientist

Mamta R Kanvinde

*****.********@*****.*** 404-***-**** www.linkedin.com/in/mamtakanvinde

SUMMARY

Aspiring Data Scientist pursuing second Master’s degree. Actively looking for full time opportunities in related fields

Knowledge of Predictive Modeling, Probability and Statistics, Machine Learning, Bayesian Networks, Python and SQL EDUCAT I ON

ILLINOIS INSTITUTE OF TECHNOLOGY - MS in Data Science; Chicago, IL May 2016

UNIVERSITY OF MUMBAI - MS in Computer Science; Mumbai, India May 2014

UNIVERSITY OF MUMBAI - BS in Computer Science; Mumbai, India May 2012 SK I L L S AND COUR S E S

Related Courses Machine Learning, Data Preparation& Analysis, Probabilistic Graphical Models, Applied Statistics, Social Network Analysis, Computational Math, Advanced Database Organization, Public Engagement and Project Management, Data Structures & Algorithms, Artificial Intelligence, Data Mining.

Programming skills Python(numpy, scipy, scikit-learn), R, SQL, weka, MS-Excel(VBA).Net, Matlab, Advanced Java, HTML

Tools MS Office, MS Excel(Advanced), R-Studio, Git-Hub, Bit-Bucket, latex, Ipython Notebook SUMMER I NTERNSH I P AS DATA SC I ENT I ST

Illinois Tech Undergraduate Admissions Department Implementation of machine learning algorithms to predict prospective students (python, weka ) May-Oct 2015

- Analyzed source dataset and implemented variable elimination by identifying co-linearity and complexity of variables

- Executed Data Cleaning and Transformation procedures on approximately 1 million records (Chi-Squared tests)

- Applied ML algorithms like Logistic Regression and naïve Bayes and validated results by delivering 75% precision

- Improved performance of the model by 6% with Support Vector Machines ACADEMI C PRO J ECT S

Study of discriminative learning techniques in ML - Logistic Regression and Neural Networks (python) Apr 2016

- Implemented Logistic Regression classifier and related activation functions with 60% accuracy and 65% precision

- Computed parameters with stochastic gradient descent & implemented Feed Forward Neural Network with 1 hidden layer

- Calculated likelihood with Gaussian Discriminant Analysis(GDA) to determine the classes and increased performance by 18%

Study of Generative learning techniques in Machine Learning using Bayes Rule – Classification (python) Mar 2016

- Implemented parameter estimation formula and computed unknown parameters using iterative approach

- Implemented Gaussian Discriminant Analysis using Bayes Rule on different distributions like Bernoulli and Binomial distr.

- Tested the performance on new dataset with cross validation to get accuracy of 65% and visualized precision-recall curve

Study of Regression techniques in ML using Numerical and Kernel Methods - Parametric Regression (python) Feb 2016

- Implemented multivariate linear and non linear models with regularization

- Computed unknown parameters with gradient descent approach and computed cost functions to get 82% precision

- Applied kernel dual solutions and tested results to get a boost of 7% in overall performance

Recommendation System for Amazon’s product dataset (python) Jan-Apr 2016

- Generated a utility matrix by filtering top rated items from a 2GB large dataset

- Implemented content & collaborative filtering algorithms to generate top recommendations for a given user

- Evaluated performances of both the algorithms with MAE,RMSE, precision and recall

Natural language processing on captions of Instagram posts to predict the activities of a user (python) Oct-Dec 2015

- Executed data collection scripts using Instagram web API in python and collected 1 Gb of data

- Implemented Natural Language Processing methods such as sentiment analysis & implemented Generative learning techniques

- Performed community detection and link prediction to get a set of users who are posting about similar content.

Sales Prediction for a retail store (R programming) Oct-Dec 2015

- Analyzed the trends in dataset with the help of visualization libraries in R

- Performed feature selections and created a time series regression model by capturing interactions between predictors

- Performed forecasting on test data with time series model to predict sales in future

Document Classification task to predict and apply labels to each document (python) Jan-Apr 2015

- Built a graphical Bayesian model to track cited documents and used the graph to create aggregations like frequency counts

- Applied logistic regression and support vector machine algorithms to classify the documents into a set of predefined labels

- Evaluated model to get 65% accuracy and improved efficiency up to 20% by using Iterative Classification Algorithm (ICA)

Web Application Development (ASP.Net) Jan-Apr 2012

- Designed and developed a photo blogging website using ASP.NET development platform with SQL backend

- Created a GUI for adding, updating and deleting photos and posts for a set of registered users and designed feedback module LEADER SH I P ROLE S

Event Organizer for college festival

- Responsible for coordinating, organizing and administering technical event of the college 2014

Marketing Head for undergraduate college festival collecting a total of INR 50 K from 5 sponsors

- Led a team of 15 marketing volunteers to execute marketing and publicity campaigns and get sponsorships 2013

Contact this candidate