Designed and developed a Python program to retrieve, clean, and summarize 1,100+ daily reports from MISO API to analyze Minnesota Hub day-ahead market energy prices into excel worksheet, increasing productivity across the team

Created server availability dashboard that linked to internal APIs and fetched data from Microsoft SQL Server using PowerBI, assisting sales reps to monitor 150+ subscribers’ EVs charging usage and identify opportunities for upselling and retention.

Analyzed and implemented a logistic regression model using scikit-learn package to predict the likelihood of peak events during summer months with 90% accuracy, supporting the programs deploying decision

Provided different data mining/cleaning/visualization requests weekly for managers and cross-functional groups to help form the design of programs/roadmaps.

Skills: Python (NumPy, Pandas, Seaborn, Matplotlib, Scikit-Learn, Plotly) , SQL, PowerBI, Excel

Worked with a team of undergraduate students and a professor on various research projects related to differrent Machine Learning topics

Evaluated the performance of  KNN, Random Forest, LightGBM, XGBoost, VecStack, and CNNs to predict housing prices in Beijing, achieving the best error rate of 16% using VecStack

Compared and analyzed the effectiveness of GloVe, FastText, and Word2Vec pretrained embedding models to build an text encoder for stock related tweet data, finding GloVe to be the best method

Contributed to 2 research papers:

• Housing price prediction via improved machine learning techniques 

• StTime-Net: Combining both Historical and Textual Factors for Stock Movement Prediction.

Developed Scala scripts in Spark to extract, mine, and analyze clustered URL access data stored in the Hadoop Distributed File System, uncovering trends and patterns in web traffic to learn unique features of cyber security attacks.

Generated a tree-based flowchart to predict potential malware attacks, enhancing the feature engineering process of building Machine Learning models that can prevent cyber security attacks targeting the organization.

Attended SQL training to help handling database

Analyzed old source code of the invoice websites and help team improve the website

Learned HTML, CSS, JSON