Michael Eryan

Exploring variables and relationships using summary statistics and plots.

Languages: Python2, R and SAS

Summary statistics and plots using an experiment with playing cards. Demonstration of the Central Limit Theorem.

Language: Python2

Hypothesis testing, T-test using data from a "Stroop" experiment.

Language: R

Data wrangling, cleaning and analyzing in MongoDB.

Language: Python2 and MongoDB

Intro Machine learning: identifying persons of interest using Enron employee financial and e-mail data.

Language: Python2

A/B testing to find out the effectiveness of a new website feature.

Language: R

Power Analysis is an interactive web application that is built on top of the "pwr" R package and allows to estimate the required sample size for a T-test given the effect size, confidence and power.

Language: R and Shiny

An online presentation explaining how to calculate the effect size statistic needed as an input for the Power Analysis application.

Language: ioslides and Rpubs

A dashboard showing an overview of the Prosper loan data.

Language: Python2 and Plotly

A Tableau story with a thorough exploration of Prosper Loan Data.

Language: Python2 and Tableau

A simple web application that allows users to analyze movie review sentiments. The ML algorithm guesses whether the review is positve or negative and provides the probability as well. The users have the option of submitting feedback to help improve the model in real time.

Language: Python3, scikit-learn and Flask web app

Ad-hoc functions in Python3.

Language: Python3

Supervised model evaluation and validation for Boston housing prices.

Language: Python3, scikit-learn

Supervised/predictive learning to finding donors for CharityML.

Language: Python3, scikit-learn

Unsupervised learning to creating customer segments.

Language: Python3, scikit-learn

Convolutional neural network to identify dog breeds.

Language: Python3, Keras

Reinforcement learning to training a quadcopter drone to fly.

Language: Python3, Keras

Generative Deep Learning using recurrent neural network to create new TV scripts.

Language: Python3, Keras