Early stopping with reinforcement learning

Build and benchmarking of DDQN for early classification of time series.

Go to project GitLab

Python Pytorch Reinforcement Learning Time Series Early Stopping Double Deep Q-Network

Thompson Sampling with Tensorflow Probability

Run through of Thompson sampling for several different types of multi-armed bandits.

Go to project GitLab

Python Tensorflow Probability Reinforcement Learning Multi-armed bandits

Credit Card Fraud Detection

Use obsfucated transaction data to develop a model for detecting fraudulent credit card transactions

Go to project GitLab

Classification Sklearn SVMs hyperparameter optimization

Introduction to Statistical Learning Solutions (Python)

Problem solutions for G. James et al.'s An Introduction to Statistical Learning. Where the textbook uses the `R` programming language, these solutions all use python with popular statistical and ML modules.

Go to project GitLab

Regression Classification Sklearn pandas

Disaster tweets

Create a model to determine whether a tweet is about a disaster or not.

Go to project GitLab

NLP BoW Classification spaCy Tweets

Ames house price predictions

Predict the price of houses using a variety of models. Determine the best model for this approach.

Go to project GitLab

XGBoost Feat. Eng. Pipelines Regression Rand. Forests

Other Projects

Local pySpark Setup

Local setup of pyspark on linux virtual environment for development and learning.

Go to project

Linux Virtual Env Spark pySpark Jupyter