TAMOGHNA CHATTOPADHYAY

Data Analyst with * year of experience executing data driven

solutions to increase efficiency, accuracy and utility of internal data processing. Experienced in creating data regression models and using predictive data modeling to deliver insights. CONTACT:

Email: adjktz@r.postjobfree.com

Phone: +1-213-***-****

LinkedIn: linkedin.com/in/tamoghna7/

GitHub: github.com/TamoghnaChattop

EDUCATION:

University of Southern California

MS ELECTRICAL ENGINEERING 2019

Gujarat Technological University

BE ELECTRONICS ENGINEERING 2017

Kansas State University

SUMMER EXCHANGE STUDENT 2016

GRADUATE COURSEWORK:

Probability for Electrical Engineers, Linear Algebra for Engineering, Mathematical Pattern Recognition, Digital Image Processing, Optimization for Information and Data Science, Machine Learning from Signals, Simulation Methods for Stochastic Systems, Deep Learning Directed Research

SQL for Data Analysis MODE, UDACITY

Deep Learning Specialization Coursera

Tableau Training: Master Tableau for Data Science Udemy SKILLS:

Programming Language Python, C++, MATLAB, C, SQL, HTML Libraries / Framework NumPy, sklearn TensorFlow, Keras, OpenCV, Pandas, Tableau, Hadoop, Spark

Applications Microsoft Office tools – Excel, Word, PPT Algorithms Logistic Regression, K-Nearest Neighbors, Decision Tree Random Forest, Support Vector Machines, Linear Regression, Lasso Regression etc.

POSITION OF IMPORTANCE:

Senator (EE Department) – Viterbi Graduate Student Associ. Part of Graduate Student Government at Viterbi School of Engineering, USC focused on student issues and engagement

STUDENT COMPETITIONS:

USC DATA ANALYTICS COMPETITION SEPT 2018

Won first place in the competition organized by USC Hospitality and Applied Statistics Club. Used Python and Tableau to analyze market share, YOY growth and popularity assessment of product. Also devised the Growth Share Matrix for it.

WORK EXPERIENCE:

DATA SCIENTIST CODERSDATA June 2020 –

• Working on inferring insights from customer data to develop marketing strategies. Using statistical techniques for hypothesis testing to validate data and present conclusions to team.

SOFTWARE DEVELOPER USC GAMEPIPE LAB June 2019 –June 2020

• Worked on implementation of image recognition, segmentation and machine learning in real world games. Used Python, Keras and TensorFlow to code the segmentation algorithm.

RESEARCH EXPERIENCE:

RESEARCH ASSISTANT USC Jan 2019 – May 2019

• Worked on finding a correlation between MRI scans and Fluid intelligence scores using neural networks. Got a MSE of 0.70 on Test Dataset.

RESEARCH ASSISTANT LA CHILDREN’S HOSP. Sept 2017 – Dec 2017

• Worked on Python script to do segmentation of neonatal brains into gray matter and white matter adopting Diffusion Tensor Imaging. PROJECTS:

Mars Orbiter Landing Using Reinforcement Learning April 2019

• Constructed and registered a new Open AI gym environment based on Lunar Lander v2

• Applied RL algorithms like Deep Q-Learning, Policy Gradient and Asynchronous Actor Critic to solve the new environment. House Price Prediction – Airbnb Dataset Dec 2018

• Analyzed the price prediction capability of different algorithms like linear regression, tree-based regression, lasso and ridge regression along with attributes like sparsity and algorithmic complexity for Airbnb house price listing dataset.

• Performance of each model was evaluated using four different metrics – R2 value, mean squared error, mean absolute error and median absolute error. We got the best results for Gradient Boosting regressor. Face and Object Detection & GANs Aug 2018

• Created a python script to apply Viola Jones algorithm. Further carried out Eye and Smile detection applying Haar Cascades on both still pictures and real time Video through webcam.

• Executed Object detection through Single Shot Multi Box Detection algorithm on small videos

• Analyzed and performed basic image generation employing Generative Adversarial Networks

Image Processing Algorithm Project May 2018

• Developed code for a library of functions in C++ for performing homography, morphological transforms, image warping and image segmentation

• Tested SAAK Transform as an alternative to CNN for combining image understanding and compression pipeline. Utilized Python and OpenCV to implement SIFT, SURF algorithms and SAAK

• Realized Convolutional Neural Network based on LeNet-5 architecture on MNIST database. A test accuracy of 99.28% was achieved on dataset Online News Popularity Dataset Analysis May 2018

• Created a pattern recognition system to determine popularity of a Mashable News article

• Performed PCA for dimensionality reduction and compared its result with feature selection done by a neural network

• Applied several ML algorithms (SVM, NN, Bayes, Random Forest) to make predictions and achieved best accuracy of 69% and an F1 score of 0.67 for multiclass SVM Classifier

Contact this candidate