KHUSHBOO PATEL

San Francisco adlfay@r.postjobfree.com 628-***-**** linkedin.com/in/Khushboo-patel1 kpatel47.wixsite.com/website-1 EDUCATION Master of Science - Data Science, University of the Pacific, San Francisco, USA (GPA: 3.71/4) 2019 - 2021 Bachelor of Technology - Instrumentation & Control Engineering, Nirma University, India. (GPA: 7.93/10) 2015 - 2019 Relevant Courses: Machine Learning, Neural Networks, Analytics Computing, RDBMS, Data Wrangling, Time Series Analysis, Data Engineering, sdFrequentist Statistics, Bayesian Statistics, Linear Algebra, NoSQL, Data Visualization, Analytics storytelling, Customer Analytics, fraud detection. SKILLS

• Languages/ Tools: R/ Rstudio, SAS, Python, SQL, C, C++, HTML, OpenCV, Matplotlib, Jupyter Notebook, Google Colab

• Frameworks: NumPy, Pandas, SciPy, NLTK, Scikit-learn, TensorFlow, Keras, PyTorch, NLTK, streamlit, Matplotlib

• Techniques: A/B testing, ETL, Data science pipeline (cleansing, wrangling, data visualization, modeling, interpretation), Statistics, Time series, Hypothesis Testing

• Business Intelligence: Tableau, Power BI, Stata, MS Excel INTERNSHIPS

• Database: MSSQL, MySQL, MongoDB, Cassandra, redis Machine Learning Intern Sunflowee Biztech Private Limited May 2020 - Aug 2020

• Developed an app for real time face detection from CCTV feed for person counter using OpenCV in Google Colab with GPU runtime by making the process faster by 30% by adjusting confidence threshold and accuracy up to 76%.

• Implemented deep learning model (YOLO) model for object detection and Caffemodel for gender + age group prediction using CNN architecture.

• Added option to browse and upload the file from device or directly paste from YouTube link (using pafy) and download process video and csv file. Built interactive analytics dashboard with visualizations (plots) using streamlit. Data Analyst Intern Moodcafe Wellness Solutions Private Limited Dec 2018 - Aug 2019

• Designed surveys to study the mental health awareness in India. Drafted questionnaire, collected data, cleaned and formatted responses using SAS.

• Identified current problems, awareness and accessibility of mental health, target groups, market potential and gaps in existing solution.

• Co-presented business idea at Startathon: Inter-college business-plan competition at IIM Ahmedabad and won 1st prize. Research Intern Siemens June 2018 - July 2018

• Designed a prototype in PLC to reduce the cost of turbine lube oil by 40% by proposing a new cycle air preparation unit design.

• Enhanced the logic design of communication protocol model for DCS system and field devices along with a comparison sheet of different protocols.

ACADEMIC Resume Ranking PROJECTS system using NLP Python, NLP

• Accomplished data cleaning by removing punctuations and stop words, tokenizing the words and vectorizing the data.

• Added a new feature (length of text and percentage of punctuations in text) with a hypothesis for higher accuracy in classification. Coupon Recommendation System Python, sklearn (clustering technique)

• Created recommendation engine for email coupon campaign for the university canteen where students will be given coupons based on their characteristics and food habits.

• Predicted the food liking of a student using data mining tools by clustering data by kNN algorithm using sklearn. Lift management system C++

• Developed and simulated an integrated lift management system using C++ to run 5 parallel lifts which reduces travelling time by ~30%. Face detection and filtration Python, OpenCV, Tensorflow

• Developed code for facial feature detection, representing face as a set of measurements and encoding faces.

• Designed a face recognition system that can detect, identify the faces and even modify using “digital makeup” with accuracy ~89%. Personality prediction based on phone usage R Studio, PCA, SVM

• Reduced dataset dimension by analyzing collinearity using principal Component Analysis (PCA) and Singular Value Decomposition (SVM).

• Analyzed relationship between cell phone usage and personality of the user by designing surveys, collected and cleaned data along with creating data regression equation.

Credit Risk Analysis R, MS Excel

• Performed data cleaning, feature engineering and designed a predictive solution to identify credit defaulters using data minings and machine learning algorithm like PCA, Decision Trees, Random Forests, Gradient Boosted trees, and SVM. Identified credit defaulters with 86% prediction accuracy.

PUBLICATION Comparative analysis of different methods for fruit quality detection (IEEE - CCECE 2019)

• Analyzed different methods including machine learning technique to determine fruit quality based on accuracy in prediction, time taken for the process, type of data used and computational complexity.

Contact this candidate