↑ Click the title for github repository!
- Using artificial neural networks, created >90% accurate diagnosis for Parkinson’s Disease, comparing it to XGBoost ML
- Includes Visualizations and finds correlations within the data
- Skills Used: Tensorflow/Keras Deep Learning Nerual Networks, Sklearn, XGBoost Classifier

↑ Click the title for github repository!
- Using two Kaggle datasets (Glassdoor) with information (education, experience, etc.) to predict annual salary
- After Data Preprocessing, EDA, and visualizations to understand dataset, tested Regression Algorithms for optimization
- Two files comparing the difference of accuracy, NLP Count Vectorizer vs simple One Hot Encoding
- >95% accuracy using NLP and Count Vectorizer and tweaking using GridSearchCV
- Skills Used: NLP (Count Vectorizer), ML (Decision Tree Regressor), Preprocessing, EDA

↑ Click the title see Deepnote!
- Hypothesis tests on correlations between excess noise and Population Density, using correlations and histograms
- Based on database of major cities in the US and their noise pollution created visualizations to predict excess noise based on variables
- Skills Used: Machine Learning w/ Sklearn, Hypothesis Testing, EDA