Joe Dobrow

Intro

I am a highly motivated and results-oriented data scientist with five years of professional experience within university study and public math education. I gained industry relevant skills in machine learning through Thinkful’s Data Science Flex program and working with a mentor fluent in professional standards. Looking to secure a position in an organisation where I can bring immediate and strategic value for the business with my expertise in data, statistics, and machine learning. Check out my work.

Work

Book recommender using NLP, SVD, K-Means clustering, web scraping, and more! I improve upon existing baselines using original strategies and novel data utilization
Book Recommender

Predicting insurance fraud using neural networks and unsupervised clustering. I use oversampling techniques to deal with imbalanced data, and show why accuracy is a bad metric of success for this model.
Predicting Fraud with Clustering

Learning to classify handwritten digits without neural networks. This project highlights my understanding of computer vision and feature engineering by coding from scratch with 96% accuracy.
Computer Vision from Scratch

Competitor on Kaggle.
Profile

About

Years of experience in customer service, teaching, and working on teams with all kinds of human beings. I have a total fascination with developing AI, which led me to study machine learning and data science. Right now, I am focused on natural language processing using recurrent neural networks, and want to bring my passion and talent for problem solving to real-world applications of AI.

Contact

• fa-linkedin
• GitHub

jdobrow@gmail.com

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';