Ash Mundhe
*******@*******.*** 201-***-**** Brooklyn, NY Open to Relocation LinkedIn Github Do Not Require Sponsorship To Start Work
EDUCATION
Drexel University, MS Data Science (STEM, Dean’s Scholarship, GHC, RTC) PA, USA - June 2020 Maharashtra Institute of Technology, BS Computer Science and Engineering (First Class, HeForShe) MH, India - May 2018 WORK EXPERIENCE
Data Analyst March 2021 - Pres
Lakeshore Global Remote
● Automated the process of loading data from different sources and merging it to clean and pre-process it in Python
● Performed time-series analysis on IoT system's data to predict hourly electrical consumption in commercial buildings
● Analyzed data to find environmental variables that influence the energy consumption Data Science Research Assistant - Visual Analytic Observatory April 2020 – Dec 2020 Drexel University Philadelphia, USA
● Feature engineered a comprehensive dataset of over 150k COVID-19 scholarly articles by automating web scrapers.
● Developed and deployed predictive models identifying the potential of new research articles using SVA.
● Identified predominant subjects using undirected graphs, centrality and other Network Analysis techniques in Python.
● Created visualizations representing concentration of themes based on the references for impactful visual accessibility for researchers in Citespace.
PROJECTS
Credit - Card Approval (Major Credit Card Company’s Data) Feature Engineering, Model Ensemble, Parameter Tuning
● Predicted the approval rate of credit cards for clients with sub optimal credit scores.
● Developed a novel ensemble feature selection algorithm to overcome the curse of dimensionality combining RF, Lasso Regression and PCA on the 1200 attributes of quantitative data.
● Achieved classification of clients with F1 score of 64% leveraging logistic regression. Market Basket Analysis - Python, Data Mining, Marketing Strategy, Consumer Data, Clustering, Customer Segmentation
● Discovered trends and patterns of customers’ buying habits by generating association rules from transaction data.
● Designed smart marketing strategies by identifying key itemsets using Apriori Algorithm.
● Achieved and improved recommendation platform by customer segmentation into clusters using K-means and PCA. Fake News Detection - PySpark, NLP, Optimization, GCP, Classification, Unstructured Data, TF-IDF, Optimization
● Devised text classification scripts in Pyspark to create n-grams and compute TF-IDF on highly-read articles in the USA.
● Achieved an improvement of 4% in a state of the art published study by developing and comparing classifier algorithms like LSVM, Naive Bayes and Decision Trees.
● Deployed the best model on GCP architecture with LSVM giving a high ROC AUC of 94% detecting the fake articles. Tweefy - Python, NLP, Sentiment Analysis, Data Cleaning, POS Tagging, Tokenization, Web Scrapers
● Retrieved twitter user demographics and music streaming behaviour using API calls and self devised scarpers.
● Analyzed the polarity and subjectivity of tweets by performing sentiment analysis using TextBlob.
● Recommended mood enhancing songs by analyzing the relationships among valence of tweets, mood and linked musical media - audio features to recommend songs to improve user’s mood. SKILLS
Programming Languages: Advanced Python (NumPy, Pandas, Scikit-learn, Re, SciPy, NLTK), Adv SQL(Oracle, MySql), NoSql(MongoDB, Cassandra), Excel, R, C++
Machine Learning: Regression, SVM, Naive Bayes, Random Forest, Gradient Boosting, KNN, PCA, Deep Learning Data Analysis: NLP, Statistical Modeling, Predictive Modeling, Data Mining, Data Wrangling, Feature Engineering Visualization Tools: Tableau, Power BI, Looker
Marketing Analytics: RFM Analysis, Pricing Value, Customer Lifetime Value, Targeting Strategies, Google Analytics CONFERENCES AND ACTIVITIES
GHC ‘20, Global AI Hackathon, NLP Summit 2020, DataCated Conference 2020, International Conference of IoT
Leadership : Led the creative and marketing team for HeForShe campaign, Head of Tech at Texephyr 2015