Ming Lyu Email & Phone Number

美国 加利福尼亚州 圣莫妮卡

• Implemented unsupervised topic modeling on articles from 20+ websites using an approach based on TF-IDF scores from n-gram frequencies, indexing scores, and other relevant metrics for internal linking to increase back-links.• Utilized sentence-transformer for efficient article clustering based on predefined topics, generating sentence embeddings similar to word2vec.• Leveraged LLM(large language models) like ChatGPT for accurate topic tagging across diverse articles with appropriate prompts, enhancing content categorization and accessibility and conducting in-depth performance and competitor analysis, translating data into actionable insights for strategic decision-making.• Designed and implemented multiple NLP-related widget applications, then containerized them using Docker for efficient deployment. These applications were deployed on Nomad servers, serving as API endpoints, which enhanced website functionality and contributed to increased web traffic by 15%.• Developed data visualizations and dashboards with Matplotlib for conveying insights and actionable recommendations. This approach led to the removal of 25% of low-quality pages, which only contributed 5% to the overall domain revenue

美国 加利福尼亚州 洛杉矶

Santa Monica, California, United States

• Analyzed our competitors’ websites by assessing their keyword traffic and website features, in order to improve our search engine optimization (SEO) strategy through a comprehensive gap analysis.• Developed and optimized an algorithm capable of identifying suitable rhyme words for a database of 70,000 synonyms, enhancing the user experience and promoting greater creativity in language use.• Created a database of 70,000 synonyms with matching sentence examples based on part-of-speech (POS) tags and readability scores, along with famous quotes for context and inspiration.• Clustered Google’s n-gram data using sentence embedding, cosine similarity, and hierarchical clustering to build synonyms for phrases (containing at least two words) to extend the 70k synonyms database.

Los Angeles, California, United States

• Designed a scalable experimental structure for testing various deep learning and machine learning models.• Utilized transfer learning techniques with popular models such as VGG19, ResNet50, and Inception V3 to accurately classify microstructure images, achieving a validation accuracy of approximately 0.8.• Conducted visualization of images within the convolutional neural network to analyze feature extraction and model interpretation.• Worked in collaboration with the front-end team to automate the model optimization and prediction process.

Hangzhou, Zhejiang, China

• Designed and implemented a comprehensive schema and rules for feature extraction from medical record texts, specifically focusing on lung cancer cases. Analyzed a vast dataset of over 10,000 patient records to fulfill the requirements of Zhejiang Cancer Hospital, a leading institution renowned for cancer treatment in China.• Conducted a meticulous error analysis to enhance the feature extraction rules, leading to an impressive 90% accuracy rate in extracting crucial features, including lung cancer originality and the timing of the initial diagnosis.• Worked closely with a multi-disciplinary team of medical professionals, leveraging their domain expertise in lung cancer treatment, to ensure that the rules for feature extraction aligned seamlessly with medical standards and best practices.

Terre Haute, Indiana Area

• Conducted thorough and in-depth research on cryptocurrency and blockchain technologies, exploring their underlying principles and applications within the financial and technological landscape.• Provided comprehensive overviews of prominent cryptocurrencies, including Bitcoin, Ethereum, and IOTA. Performed detailed analyses of each cryptocurrency's design and architecture to assess their real-world potential and their ability to deliver on their promises.• Led research efforts focused on applying regulations to the cryptocurrency industry, with a particular emphasis on combating issues related to money laundering and illicit activities.• Demonstrated a keen interest in the convergence of blockchain techniques with the Internet of Things (IoT), with a strong fascination for IOTA's innovative tangle system and its potential to revolutionize the IoT ecosystem.

Master Of Science - Ms, Computer Science

Activities and Societies: GRIDS (Graduates Rising in Information and Data Science)

Bachelor Of Science - Bs, Double Major, Applied Mathematics And Computer Science

Bachelor of Science degree with double major in Mathematics with a concentration in discrete applied math and Computer Science