Data Scientist

MARCELO FUENTES DATA SCIENTIST

SUMMARY

SKILLS & INTERESTS

LANGUAGES: Python (Pandas, NumPy, Scikit-Learn, Matplotlib, Seaborn, Plotly), SQL, VBA MACHINE LEARNING: Hypothesis Testing, Linear Regression, Logistic Regression, Vector Autoregression, Auto Regressive Integrated Moving Average Regression, Ridge Regression, k-Nearest Neighbors, Random Forest, Naive Bayes, K-Means Clustering, Neural Networks, Natural Language Processing, Keras, Tensorflow, Sequential, pySpark, Spark INTERESTS: hiking, biking, skateboarding, cooking, baking, traveling EXPERIENCE

Mar ’18 - Present Associate Manager Risk Analytics (Transferred from Regulatory Reporting), SS&C GlobeOp Maintains and develops statistical and analytical tools using MS SQL Server, MS Excel, MS Access to enhance the risk management tool kit used for risk monitoring purposes at the security and portfolio level. Reconciles holdings, exposure and transactions databases. Performs stress testing, scenario modeling, historical and parametric Monte Carlo simulations. Supports and updates the SQL Risk Management database to help with risk analytics execution and risk reporting Jan ’17 - Mar ‘18 Analyst Organizational & People Analytics, Aon Hewitt Managed big datasets for asset management firms and generated insightful visualizations to report on their financial performance, pay practices, goal setting, etc., using MS Excel and MS Access. Re-purposed datasets by developing VBA macros streamlining data cleaning, data management & report generation. DATA SCIENCE PROJECTS

S&P 500 Returns Forecasting

Python tools: Scipy.stats, Scikit-Learn, Statsmodels, VAR, ARIMA, Ridge Regression, Recursive Neural Network Summary: performed time series analysis using daily opening and closing prices for the S&P 500 Index and its top five constituents (AAPL, AMZN, FB, GOOG, MSFT), used three regression models and a recursive neural network to forecast a month of daily closing prices. The best model had a mean absolute percentage error of 0.4% (Ridge Regression). Spotify Song Popularity Predictor

Python tools: Scipy.stats, Scikit-Learn, Statsmodels, Random Forest, k-Nearest Neighbors, Logistic Regression, OLS Summary: used classification algorithms and song demographics such as loudness, acousticness, duration, etc. to train classification algorithms into predicting songs’ popularities on a 1-5 scale. This dataset came from Spotify and is comprised of 160K+ songs. The best model was a logistic regression model with 84% accuracy. EDUCATION & CERTIFICATIONS

Springboard - Data Science & Machine Learning Track - Feb 2021 500+ hour certification, covering data extraction, analytics, machine learning, etc. using Python and SQL Charter Financial Analyst (CFA) Level III Candidate - June 2021 Covered topics ranging from investments and economics to probability and statistics Syracuse University - May 2016

Bachelor of Science in Economics & International Relations, Minors in Mathematics & Finance mm.fuenteslopez@gmail 315-***-**** Brooklyn, NY marcelo-fuentes-lopez mmfuente95 Data Scientist and Risk Analyst with a Bachelor’s degree in Economics and 4 + years of experience in financial instrument pricing and data analytics. Technical expertise using Python, SQL, MS Excel and MS Access. Keen interest in statistical analysis, machine learning and process optimization. Specialized in ETL, data analytics, data visualizations, report generation and financial instrument analysis