Hi it's Oreolorun!

About

As a Civil Engineering graduate (Transportation Engineering specialisation) I had been exposed to the possible benefits of self driving vehicles in transportation corridors. However, something unlikely happened as I became keenly curious about the technology that made them possible. This curiousity drove me to machine learning and over the past few months I've taken a deep dive into the field and decided to turn the source of my curiousity into a career.

I am a Machine Learning Engineer well grounded in building machine learning and deep learning models with Python. I am proficient in the use of Pandas for working with tabular data and Numpy for scientific computing, I am also well versed in the use of Scikit-Learn and PyTorch for building models for structured and unstructured data respectively.

I am skilled at building web scrapers for data collection and curating my own datasets. In addition to this, I am profient at writing Flask web APIs for model deployment, having deployed models on Streamlit cloud and Heroku. I am experinced in version control using Git and containerisation using Docker. In my free time I love to write vanilla code for image processing, machine learning and deep learning concepts then proceed to blog about them.

Resume (360kb)

Projects

Image Recognition

An image classification project where I utilized an image dataset I had personally scraped and curated in building a deep learning model capable of classifying cars as either Sedan, Coupe, SUV or Truck. I used a custom built Convolutional Neural Network architecture and I was able to achieve 95% classification accuracy which rose to 96% upon ensembling.

Libraries/Tools: OpenCV, PyTorch, Numpy, Shap, Flask, Git, Docker, Heroku, Streamlit

Github | Streamlit App | Flask App | Docker Image

Visual Similarity Recommendation

A computer vision project where I built a web app capable of recieving an image via upload, looking through a database of images and returning images most similar to the uploaded image. This web app is powered by a neural network which serves as feature extractor and cosine similarity for computing similarity scores.

Libraries/Tools: OpenCV, PyTorch, Numpy, Shap, Flask, Git, Streamlit

Github | Streamlit App | Article

Customer Churn Classification

A binary classification project where I built and optimized a tree based model capable of predicting if a bank customer is likely to terminate his/her account based on attributes about said customer and their bank account usage. Model explainability is built into the deployed web application for each classification.

Libraries/Tools: Scikit-Learn, Pandas, Numpy, Seaborn, Matplotlib, Shap, Streamlit

Github | Streamlit App

Web Scraping

A data collection project where I built a web scraper which is capable of scraping images with in-built data cleaning features such that it checks for and deletes duplicated as well as non-pertinent instances of images. Using this scraper I built and curated the image dataset which I used for the Image Recognition and Visual Similarity projects.

Libraries/Tools: BeautifulSoup, Requests, OS

Github

Skills

Skills

Regression, Classification, Image Processing, Engineering, Programming, Mathematics, Data Collection.

Technologies

Python, Numpy, Pandas, Scikit-Learn, Pytorch, OpenCV, Flask, Git, GitHub, Docker, BeautifulSoup, Streamlit, Jupyter Notebook, Pycharm, Heroku.

Contact

• Twitter
• LinkedIn
• GitHub
• Blog

Elements

i = 0;

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

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