Projects
Here are some of my open-sourced history projects
Recent projects are currently priviate
Financial News Sentiment Analysis System based on BERT
[github] [document] [video]
• Developed Python web scrapers to automatically extract financial news from multiple platforms and label the text as Positive, Neutral, or Negative.
• Built and fine-tuned a classification network based on BERT(pre-trained model) with accuracy over 80% using Keras framwork
• Deployed a visualization website to display model classification results interactively.
Indoor HIIT Motion Recognition
[github] [document] [video]
10+ stars on Github This notebook introduces the development and deployment process of a fitness motion recognition algorithm based on built-in smartphone sensors. The algorithm identifies the user’s current action using data collected from the smartphone’s accelerometer and gyroscope sensors. For deployment, TensorFlow Serving was used to deploy the model to a server, and a Flask web framework was used to build an API service interface, allowing WeChat mini-programs to call and return prediction results.
MachineLearning with C
[github] [document] [video]
Those Project are written in complete C, implement some famous Machine Learning Algorithms