My Portfolio
Resume
Professional Skills and Capabilities
- SQL, Python (NumPy, pandas, scikit-learn, seaborn), Tableau, Excel
- Data Science, Data Engineering, Data Analysis
- Data Cleaning, EDA, Data Visualizations
- A/B Testing, Segmentation Analysis, Multi-touch Attribution Analysis
- Marketing Analytics, Consumer Insights, Growth, Strategy, Acquisition
- Google Suite and Microsoft Suite
- Communication, Storytelling, Presenting, Public Speaking
- Leadership, Adaptability, Curiosity, Creativity
Projects and Articles
Click the Images below to redirect to my full length write-ups
Vaccination and Severe COVID-19 Infection in Israel Post Delta Variant
In this paper, I investigate whether the effectiveness of COVID-19 vaccines has changed following the emergence of the Delta Variant. Utilizing data from Israeli Public Health, I calculate age-adjusted efficacy rates and confidence intervals for both vaccinated and unvaccinated populations. Ultimately, my analysis reveals a decrease in efficacy rates from over 90% to 88.9% and highlights the impact of age on infection rates.
Predicting March Madness Games and Designing a Better Tournament
In this paper, we predict the results of NCAA WBB playoff matches and assess the structure of the current playoff format, relying solely on regular-season data. Our investigation is based on the Bradley-Terry Model, which we adapt to consider variables like game date and home-field advantage. Ultimately, we discovered that a Bayesian model incorporating home-field advantage yielded the most accurate predictions. Moreover, we determined that the prediction accuracy was only sufficient for a 32-team playoff structure.
Statistical Modeling Regarding Weight and Grocery Shopping Patterns
In this paper, we employed CDC Public Health data to investigate the key factors of both individuals' weight and their decision to purchase prepared food from grocery stores. Through extensive modeling techniques including Best Subset Selection, Random Forests, Bagging, and Logistic Regression, our aim was to identify these determinants and evaluate the reliability of our findings. While we identified numerous factors influencing weight, our attempt to profile prepared food buyers yielded inconclusive results. We were able to show which models had the best predicting power and lowest error for these specific scenarios.
Possible Gender Differences in Overconfidence Study
In this paper, I review previous research on the correlation between overconfidence and mindset types (fixed or growth). Although prior studies suggested a consistent relationship across genders, I aimed to investigate whether gender influences overall confidence levels. Through regression models and an ANOVA test, I ultimately demonstrate that males indeed exhibit elevated levels of overconfidence.
College Football Matchup Predictions
Does Music Genre Affect Memory?
In this paper, we attempt to predict College Football matchup scenarios using a Support Vector Machine model. We provide a comprehensive overview of the analysis procedure, spanning from data preprocessing and exploratory data analysis to modeling and future directions. Ultimately, we developed a model with a notably high accuracy rate of 70%.
In this paper, we reflect on an experiment we designed, and attempt to draw conclusions from its results. Our experiment investigated if specific music genres improved memory and recollection abilities. By employing various statistical techniques to evaluate our participant results, we were able to show that music genre did not have a statistical impact on memory levels.
Rudderstack Internship Experience
In this Article, I share what I learned, developed, and experienced during my 5 month internship with Rudderstack. I explore topics like digital marketing, campaign management, campaign performance tracking, and how we used Rudderstack to easily understand our marketing performance metrics.
Contact
Connect
hlassa@umich.edu
© 2024. All rights reserved.