Junlin Wang's Project Portfolio

Youtube Spam Detector

Hands the spam hammer down on comments. Given a Youtube video link to our web app, the web app will display the first 100 comments along with the probability of it being spam using an RNN. You can view the most frequent spam words in a word cloud visualization.

The RNN model is trained on Youtube Spam Collection Dataset from UCI Machine Learning Repository. The RNN model incorporates LSTM layer, Bidirectional layer, Convolutional layer, etc. Stanford's Glove word embedding was used to boost the performance of the RNN model. Trained the model using Google Compute Engine.

Airboard

Developed Air board, an Android app using ultrasonic waves transmit keypresses to smart phones.Learned and utilized Android NDK and IME for development and ultrasonic encoding methods.

Automatic Stock Trader Using Q Learning

Conducted a research report on machine learning algorithms (mainly reinforcement learning) on stock markets. Wrote and published paper of Profitability of Q-Learning Implemented with Statistical Arbitrage on U.S. Equity Market in school journals.

Pinmap - Fast Point Retrieval

Developed a technique and improved the search time of a point from O(n) to O(1) time using render to texture technique in webgl and JavaScript with Prof Zhao Shuang. Utilized the rendertotexture to do a two pass rendering.

Mandelbrot Set Explorer

Built an python application using tikinter graphics that enables the users to visualize and explore Mandelbrot Set. Used Cython to accelerate the computation.