Yitong Hu's repositories
CNN_feature_visualization_by_tensorflow
I trained a CNN to solve a particular classification problem(SVHN&MNIST), then treated input image as optimization variable and optimized for the image that maximally excites a certain layer. I optimized by regularization to find images more recognizable that excited different layers of the network.
RF-prediction-problem
Here is a prediction problem using RandomForest with feature transformation attempt and parameter tuning to improve F1&AUC successfully
spider163
抓取网易云音乐热门评论
transfer
Transfer Learning for Sequence Tagging with Hierarchical Recurrent Networks
Text-mining-project-The-use-of-personal-pronoun-in-Inaugural-Speeches
Here is a data analysis case of the general trend of presidents using personal pronouns and cluster the extreme users(top 10 using a certain personal pronoun separately at a high frequency) to find their features,then analyse based on each sentence to see the content and emotion it conveys.
RShiny-app-NYC-Perfect-Planner
In our app, after entering your current location, you can choose your top 3 “Most want-to-go places” to search for recommended areas that contain all your choices. Among all the recommendations, you can zoom in to have more detailed information about that area and then decide where to start your perfect day!
Classification-Problem-using-different-ML-method
Here is a classification problem with different ML model built for grayscale images of poodles versus images of fried chickens to get the highest accuracy.
Movie-Recommendation-Shiny-App
In this project, we build a movie recommendation App(https://adsproject5group10.shinyapps.io/moviebeats/) based on the OMBD dataset.
Name-disambiguation-in-author-citations-using-Spectral-clustering-EM-algorithm
Here is the implement of two different algorithms to deal with name disambiguation in author citations,separately used Spectral clustering & EM algorithm with C6 tau-coauthor constraints
pandas_exercises
Practice your pandas skills!
CWS_LSTM
EMNLP2015_code_Long Short-Term Memory Neural Networks for Chinese Word Segmentation