chrisshaffer Goto Github PK

covid-symptom-searches

Are Google searches for COVID-19 symptoms useful in predicting increases in COVID-19 patients? Exploratory data analysis and hypothesis testing using Google's COVID-19 Open Data database.

data_science

ds-precourse-axes

A repository to hold the challenges having to do with modifying the axes on a matplotlib object.

ds-precourse-faceting

ds-precourse-histograms

This repository holds the content for the plotting challenges in the histograms lesson.

ds-precourse-scatterplots

dsi-learn-week0-commandline

dsi_notes_strategy

A strategy for taking and using quick notes in the Galvanize DSI.

dsi_precourse

This is for practice using github/git

git-primer-checkpoint

group-playlist-recommender

Companies like Spotify create personalized recommendations. How can this be extended to groups of people? This project uses the Echo Nest Taste Profile of user play histories to create recommended playlists for groups of users.

kaggle-house-price-prediction

My submission to the Kaggle housing price prediction competition, placing in the top 1% of submissions. This uses the XGBoost Regressor, which implements gradient boosted decision trees with high speed and performance.

kaggle-titanic

My submission for the Titanic Kaggle competition with accuracy in the top 7% of submissions. Accuracy was improved by data cleaning to deal with missing values, feature engineering, one-hot encoding of categorical features, use of the XGB Classifier, and hyperparameter optimization.

nn-character-recognition

Neural network-based character recognition using MATLAB. The algorithm does not rely on external ML modules, and is rigorously defined from scratch. A report is included which explains the theory, algorithm performance comparisons, and hyperparameter optimization.

resume

sem-image-recognition

Image recognition of scanning electron microscope (SEM) images using CNNs

wafer-defect-maps

Inspection equipment for the semiconductor industry saves companies millions of dollars. This project uses the MIR-WM811K Corpus» of wafer maps to build a CNN classifier to automate classification of wafer defect patterns.