Projects

  • Scheduling-Aware Data Prefetching Based on Spark Framework
    — Proposed a scheduling-aware data prefetching mechanism. By evaluation, the mechanism reduced 14% and 17% Spark execution time with 25 GB and 50 GB datasets.
    — The prefetch mechanism was based on Spark task scheduling, data was prefetched to cache just before it will be processed.
    — The cache system was based on Alluxio, an open source in-memory distributed storage system.
    — Accepted by The 31st IEEE International Conference on Advanced Information Networking and Applications (AINA-2017).

  • FileBox
    — A Dropbox client on Windows Phone.
    FileBox is available on Microsoft Store.

  • Car Classifier
    — A Convolution Neural Network (CNN) based neural network image classification system.
    — Used Keras and TensorFlow to train the classification model; used Django and Python to develop the classifier website. Demo classifier website was deployed on Microsoft Azure.
    Demo
    ly2314/NotHotdog-Classifier
    ly2314/CarIdentifier-Web

  • PLURKLY!
    — A Plurk! client on Windows Phone.
    PLURKLY! is available on Microsoft Store.

  • Daily Pixiv
    — A Pixiv daily ranking viewer on Windows Phone.
    Daily Pixiv is available on Microsoft Store.