Marcus – Data Science, Machine Learning Engineer

About Me

Data Science and Machine Learning Engineer with experience in data processing, machine learning model development, and data transformation infrastructure. Has worked with Python, pandas, scikit learn, tensorflow, pytorch…

Experience

Data Science, Machine Learning Engineer | Data Processing

Virufy

Worked on Data Processing team to algorithmically standardize, preprocess, and predict the presence of COVID from patient data for developing artificial intelligence technology to provide free, easily accessible screening for COVID-19 from cough patterns.
Researched and identified approaches for detecting COVID coughs through machine learning, deep learning, and signal analysis using libraries PyAudioAnalysis, scikit learn, tensorflow, pytorch, scipy on 11 datasets collected by global health organizations focusing on cough audio, symptoms, and other health related features.
Built data transformation infrastructure in Python to process and feature extract from a variety of input data formats, improving the accessibility and usability of all data standardization and preprocessing scripts.
Reworked data ingest pipeline by overhauling the storage of each dataset using AWS S3 and rewriting ETL scripts.

Data Consultant | UC Berkeley Student Association for Applied Statistics

Cortex

Developed classification and machine learning models in Python libraries pandas, scikit learn, tensorflow, pytorch, to predict rehospitalization risk scores for urgent care patients achieving 78% recall, on dataset over 250,000 rows and 1000 features on a team of student consultants.
Performed data cleaning and preprocessing, feature engineering, standardizing granularity, and addressed data imbalance using SMOTE, undersampling.
Tested and tuned various models including Support Vector Machines, Decision Trees, Random Forests, Weighted Logistic Regression, Weighted CatBoost, and built prediction pipeline achieving a peak accuracy of 89%.

Skills

Data Evaluation Python pandas scikit learn tensorflow pytorch scipy Java SQL Excel Tableau Power BI Prompt Engineering matplotlib AWS S3 ETL SMOTE undersampling Support Vector Machines Decision Trees Random Forests Weighted Logistic Regression Weighted CatBoost Logistic Regression K-Nearest Neighbors XGBoost Perceptron Neural Networks Feature Engineering Data Cleaning Data Preprocessing Data Transformation Machine Learning Deep Learning Signal Analysis