to combine my love for data science/ml and the arts, i wrote a lyrics generator by artists in python using an n-gram character-based model. currently, i'm working on implementing an rnn/lstm to better capture long-term patterns.
check it out here!
to combine my love for data science/ml and the arts, i wrote a lyrics generator by artists in python using an n-gram character-based model. currently, i'm working on implementing an rnn/lstm to better capture long-term patterns.
check it out here!
key skills: natural language processing, rnn, lstm, tensorflow, python
built and optimized an autocompletion gui in java using binary search and hash list methods for my data structures and algorithms class. this gui provides quick and relevant suggestions based on a predictive text algorithm with real-time suggestions, optimizing text input and efficiency.
check it out here!
key skills: java, data structures, algorithm optimization, performance analysis, unit testing
utilized huffman coding for a compression and decompression algorithm for my data structures and algorithms class. huffman coding is a popular method used for lossless data compression, and reduces the size of input data files without losing any information, thereby optimizing storage and transmission efficiency.
check it out here!
key skills: java, data structures, unit testing, performance analysis, data compression
i trained and fine-tuned a convolutional neural network to classify clothing and accessories (fashion mnist dataset). i minimized overfitting while achieving 87% testing accuracy.
check it out here!
key skills: python, tensorflow, keras, deep learning, seaborn, matplotlib
my second-round interview to enter the duke impact investing group's data science division as an analyst required that we complete a 3-day data challenge, advising the who on which sectors to fund to improve life expectancy in developing countries. i performed statistical analyses and created data visualizations for the final deliverable slidedeck. i built a regression model and tweaked it for collinearity.
check it out here!
key skills: R, statistical analysis, ggplot, predictive modeling
working with tensorflow and cnn's for feature extraction in predicting patient outcomes.
building a pantry tracker using react.js, next.js, and firebase.