Natural Language Processing: natural langauge classification, natural language generation, machine translation, low-resource language preservation
Machine Learning: deep learning and neural networks
Theoretical Computer Science: algorithms design and complexity analysis
Other: game theroy, computational finance, data visualization
| 1. Implement Data cleaning and joining on input: data of orders and shipping costs from warehouse. |
| 2. Develop advanced Last-area-first-fit algorithm for optimizing the size of cartons |
| 3. Modify multi-objective genetic algorithms for modelling the cartons sets having the lowest predictive annual shipping cost |
| 4. Implement a VBA tool applying those algorithms |
| 5. Run input on VBA tool, and output the dimensional for each carton from the optimal carton set |
Note: If you would like to know the details about this research, please contact me via email (The research paper is almost finished). |
| Semantic Processing - Quantifier Scoping |
This project introduces a new implementation of a quanfifier scoping algorithm. This algorithm models how neurons of our working memory deal with quantifer scoping of sentences. |
| Research Poster |
| Text Mining and Data Visualization |
This project is the current research project that I am working on. We are focusing on the structural pattern within I Ching and are discovering how the gua (卦) is related to fortune telling in I Ching Corpus. |
The related contribution, "Visualizing Philosophy - I Ching" (Author: Yvette Shen, Pei-yuan Tang, and Han-wei Shen) has been accepted for RSD7 conference. |
Another related contribution, “Visualizing Philosophy: A Modern Design Approach to Understand I-Ching” (Author: Yvette Shen, Han-wei Shen, has been accepted for presentation (and long paper publication) at the Information+ 2018 conference Pei-yuan Tang). |
| I-Ching Visualization V1 |
| Income US Analysis (2016) |
This project displays and compares the expenditures among all 5 income groups in the US. |
| Multidimensional Data Visualization |
This project uses three modes of visiualization on the data points of handwirtten digits from 0 to 10. All data are composed of 784 dimensional vectors, storing all pixels values of the pictures. We reduce the data and visualize all of them in a plane (2 dimension). Hovering over the data point will display the original greyscale image (combination of its all pixel values). The "group" option forces all data points in their own clusters as concentric circles. The "TSNE" option displays all data points scattered by the TSNE dimension-reduction method, while the "PCA" option displays all data points scattered by the PCA dimension-reduction method. |
| Urban Structure |
This project is an user interactive site for discovering the cycle of buildings of Columbus, Ohio. We use all 300 thousands data to develop a user interactive website for you to view the building data. It is a group project and we have 5 people in our group. Design: Savanah, Developers: Zhenyang(35%), Me(Pei-yuan 35%), Rob (20%) and Andres (10%). We have a main page, a user interactive map page, and a data analysis page (user interactive as well). |
| Tutorial |
It is a tutorial for DBpedia that can be used for learning the relation-based database structure. |