Python with Applications Class Project - Penguins Clustering Analysis- Created a model to predict a penguin’s species based on physical attributes. Analyzed data quality, distribution, and correlation to define a hypothesis. Performed feature selection. Preprocessed data using normalization and label encoding. Deployed 3 scikit-learn machine learning models (logistic regression, decision tree, and neural network); achieved 95%accuracy from neural network model.Python with Applications Class Project - Tweets Sentiment Analysis- Identified sentiment of Twitter posts during COVID-19 pandemic. Extracted features from 44,955 tweets of unstructured data by creating a term-document frequency matrix using CountVectorizer. Categorized emotion of Twitter posts using LogisticRegression model; achieved 85% accuracy of post sentiment from unseen tweets and identified the most negative and positive words using regression coefficient.Python with Application Class Project - Colorization of Black&White Images- Prepared and trained approximately 10,000 images using a convolutional neural network combined with a classifier deployed through TensorFlow and Keras. Created local app for colorization process for any input gray images.Kaggle Guided Project - Predicting Molecular Properties- Predicted scalar constant number to generate a fast and reliable method to identify molecular properties. Ran exploratory data analysis on approximately 5 million records, distinguished different patterns of different molecule types, and created new features by identifying neighboring atoms and their distances. Built a function to automate fitting and predicting interaction between 2 atoms using Light Gradient Boosting Regressor models. Developed 8 different models for different molecule types; obtained an average log mean absolute error of -1.3.

Created metal-antibiotic complexes and compared antimicrobial activities to improve the efficiency of antibiotics. Performed analytical test of NMR and FTIR and functional test using spectrophotometer on different bacteria strains; determined that complexes of antibiotic ligands with Gallium ion are the most effective and reported results to Dr. Charles Chau.

Helped students achieve dreams of graduating college by assisting them with coursework. Collaborated with colleagues to identify student weaknesses and adjusted teaching style to meet individual needs for 15 hours a week; assisted 15-20 students per office hour session