Why do we like using Weibo? I showed some easy codes to have little analysis, which turns out to be...

U.S. general election is coming again next year. Someone uses social media as his big weapon to arouse people's attraction, which turns out to be not a bad idea. Let's go back the old days and see clearly what's happening!

If you would like to view futher of the network, here is the way:
Static Network
Temporal Network

People talks about Brexit every day. But who talks about it? How they will describe it? What is the future of Brexit? I don’t know, but I can see it from the twitter view, somehow like a God :)

Have you ever been thinking about hiring an AI to help you trade? For that purpose, I implement DDRN and DQN methods to seek money in stock market. It turns out different configurations result in differing outcome. In comparison, DQN is not as good to find the optimal Q-Table when monitoring time series, but fortunately we still have money coming by DDRN.
(PS) I will not be responsible if you lose any. But if you earn something please give me a star ;)

Thanks to Avengers series, which give us years of happiness and massive topics to discuss during social occasions or just with your girlfriend. If you are also a big fan, just go time travelling with me what audiences like to talk about the four Avengers in the last ten years, and think along what happens with you, and your life.

I implement Dynamic Programming, Monte Carlo, TD (SARSA and Q-learning) algorithms to play a game for me hereby. This might be hard to read, but the mechanisms are just so impressive. A new version of that is here:
Advance Method

I hereby compare the four optimisers during implementation of CNN. I also add the Dropout layer to see the differences. This is kind like a comprehensive review of CNN.

A partner for freshman to save money.

I hereby predict the quality of wines by differing machine learning methods, i.e. Polynomial functions and Non-Parametric Approaches. I only showed the polynomial part in the picture below, but in the essay I compared some state-of-art algorithms, e.g. SVM, Random Forest, and XGBoosting for learning and for good wines.

This is achieved by utilizing the Baidu Index, web crawler and text mining technology and implementing the combination of multivariate linear time series econometric model and BP neural network algorithm.

We buiild a search engine for literature review with lucene library, with friendly interface and functional framework. And I really hope one day it can surpass Google and Baidu.

We hereby buiild a Management Information System for a very famous restaurant in Chengdu. Please come to a visit.