Term: Fall 2019
-
Team ## Group - 6
-
Team members
- Bukhari, Syed Ahsan [email protected]
- Chen, Xiwen [email protected]
- Cho, Sung In [email protected]
- Qiu, Feng [email protected]
- Ye, Xuanhong [email protected]
-
Project summary: In this third project of GR5243 Applied Data Science, we created a classification engine for facial emotion recognition. For the feature, we used pairwise distances between all the fiducial points since using only a location of the points showed poor performance. To build advanced models compared to the given baseline model, GBM, we chose eXtreme Gradient Boosting (XGBOOST) and Support Vector Machine (SVM) as our Machine Learning technique. We used Cross Validation to tune parameters of the models, except for XGBOOST's "n_estimators" due to a high training cost of the model. The models we built clearly showed a better performance.
Contribution statement: (default) All team members contributed equally in all stages of this project. All team members approve our work presented in this GitHub repository including this contributions statement. Please find below participation of individual resources.
- Bukhari, Syed Ahsan - tuned XGBoost to improve accuracy
- Chen, Xiwen [email protected] - tuned XGBoost to improve accuracy
- Cho, Sung In [email protected] - tuned SVM and prepared final presentation
- Qiu, Feng [email protected] - tuned SVM to improve accuracy
- Ye, Xuanhong [email protected] - Wrote base script and tuned GBM, attempted CNN to do research work
Following suggestions by RICH FITZJOHN (@richfitz). This folder is orgarnized as follows.
proj/
├── lib/
├── data/
├── doc/
├── figs/
└── output/
Please see each subfolder for a README file.