🤖 Machine Learning
::Descriptions::
👁️ .:Diabetic Retinopathy
Through Kaggle's Diabetic Retinopathy Detection competition, 35,126 high-resolution retina images were used in order to create an automated analysis system that assigns a score to images rating presence of diabetic retinopathy or not. Results from the model returned 82% Training Accuracy, 80% Testing Accuracy, 88% Precision and 77% Recall. Results were very good due to image pre-processing. For patients suspected with Diabetic Retinopathy, this is implemented to provide accurate and efficient diagnosis.
.:Lung Cancer Detection
Did some data exploration and data preparation for this dataset.
.:Nuclei Viz
Detected nuclei in order to expedite cures. Nuclei were good to use because they can help detect a biological dysfunction immediately.
🚗 .:UBER
-UberX and UberXL Cancellation Predictions based on Fare/Price Setting
-UberEATS Time Series Forecasting
🏥 .:Flatiron Health
-Helping Cancer clinic understand how its cohort is given antineoplastic (anti-cancer) drugs
💻 .:Healthline Media
-Website Analytics - Traffic and Audience Analysis
💊 .:NURX
-E-commerce Telemed Conversion Funnel Analysis
🚲 .:Postmates 🌮
-New NYC Market Analysis with Geospatial Heatmap
⌚ .:Touch of Modern
-E-commerce Consumer Behavior Analysis
💹 .:6sense
-SQL queries leveraging Contact, Company and Interaction tables as well as A/B Test Experiment Design
🔨 .:Thumbtack
-Search Result Quality and Service Fee Recommendation
🚚 .:TaskRabbit
-2-sided Market Analysis and suggested improvements to increase supply and demand sides
💻 🏥 .:eHealth, Inc.
-e-commerce U.S. Health Insurance Market Analysis with focus on data wrangling and data visualization
👧 💊 🏪 💻 .:Curology
-telemedicine, personalized skin care product usage/customer retention
🏥 .:Quest Analytics
-provider directory data analysis and metrics development for accuracy reports
🏥 💻 .:Doximity
-cohort analysis and profile rank new product feature development for online networking service for medical professionals
🍆 🍐 👜 .:Instacart
-EDA (Part 1), feature engineering (Part 2), built models and model performance comparisons (Part 3)
🚚 🍎 .:Cheetah, Inc.
-Customer Segmentation analysis for Revenue Optimization (SQL and Python)
🧑🏫 .:Desmos
-Defining product launch success metrics in Ed Tech
💊 .:WithMe Health
-Data parsing on pharmacy claims data using SQL and JSON
📈 .:Short-Term Household Energy Consumption Forecasting
-Trended and partitioned data with 80/20 split to create multivariate time series model with hour time-steps. Used Facebook's Prophet to build forecasting model that produced RMSE of 0.00033
📉 .:Shipt
-Hierarchical Time-Series Forecasting for Supply and Demand