Yuchi Chen Email & Phone Number

• Hold office hours to assist students in Linux, SAS, R, Python, Spark, SQL, Parallelization, and Git

Analyzed the relationship between Twitter and cryptocurrencies, and built a return prediction model, such as ADA, BTC, ETH• Cleaned and quantified 147 million Tweet data, such as bot detection and sentiment analysis based on a hybrid of the rules-based and model-based model, and calculated PageRank scores in the retweet network to filter useless tweets• Generated an interactive dashboard to explore the correlation between attention, sentiment, and price-related features and return by Tableau and completed feature engineering using the Granger Causality Test• Built one regression model that predicts the cryptocurrency return and one binary trend classification model in python

Forecasted the country of tweets and learned user preferences by countries to improve personalized recommendation• Crawled tweets, extracted features from texts, including hashtags, mentions, shortened URL expanding, and cleaned text data by translation, tokenization, spelling check, lemmatization, removing stopwords, and punctuations by nltk• Identified distribution, trends, and relationships of features and conducted sentiment analysis of text with visualization, including wordcloud, stacked bar charts, and cat charts using seaborn• Generated unsupervised topic clustering models to analyze topical contents, including LDA (Latent Dirichlet Allocation), NMF (Non-Negative Matrix Factorization), and visualized them as an interactive dashboard by sklearn• Built a 6-country classification model using Bi-LSTM to predict country labels of tweets without location shared

Developed a data-intensive application in Cloud to recommend ERC-20 tokens on the Ethereum Blockchain• Built the combination of streaming and batch pipelines with ETL (extract, transform, load) processing using delta lake• Completed EDA to gain information related to tokens and transactions, including trends in Spark SQL on Databricks• Managed the lifecycle of ALS recommendation model by training, tuning, serving, and monitoring with MLFlow

Conducted the analysis that raising teacher pay would improve public high school academic performance• Complemented descriptive analysis to explore distributions using histograms, and box and whisker plots in Tableau• Investigated the relationships of each feature and selected features based on heatmap, joints maps, ANOVA test, Chi-Square test, component analysis (PCA), and cluster analysis (KMeans)• Built a classification XGBoost model with 93% accuracy to forecast excellent student percentages in schools

Wuhan, Hubei, China

• Assisted in cleaning and mining data of stock market using Navicat with SQL for rough stock trend prediction• Built a multiverse deep reinforcement learning algorithm with a specific environment for coupled stock trading by adjusting Gym and Pytorch• Designed and extracted features for dynamic states and partitioned reward functions of reinforcement model• Implemented Darwin Reinforcement for parallel reinforcement training solving convergence problems and parameter adjustment for a global optimization solution• Drafted compact calibration tools for mobile posture detection app

Wuhan, Hubei, China

Project: Land Use/Land Cover Classification by Combining Deep Learning and Google Earth Engine • Obtained and dealt with big geographic data and remote sensing images through GEE in Python and JavaScript• Conducted the globally applicable 100m spatial resolution land cover classification deep learning model using TensorFlow with 85% accuracy in the scope of most European areas• Obtained and analyzed dynamic ecosystem service values ranks and trends of 39 countries in Europe from 2017 to 2020, including the epidemic influence

Wuhan, Hubei, China

• Super-resolved satellite images from 60m and 20m to 10m spatial resolution by constructing two deep neural networks with TensorFlow utilizing different band information• Coordinated work distribution among group members as a group leader

Master Of Science - Ms, Data Science

Bachelor'S Degree, Remote Sensing Technology, 3.87