Datathon-techniques

Common techniques for datathon.

Visualization

• Network Yue
• Cluster
• Decision Tree / Random forest / XGBoost
• Time series
• Line plotting
• Plotting on a map

Models

• Supervised Learning techniques: https://scikit-learn.org/stable/supervised_learning.html#supervised-learning Banghua &Yiling
• Linear regression; SVM; Nearest Neighbors
• Confidence interval Yiling
• Time series Analysis Yue
• Decision Tree Yue
• Random Forest Yue
• Hypothesis testing Yiling
• Scikit-learn: https://scikit-learn.org/stable/
• XGBoost: https://xgboost.readthedocs.io/en/latest/ Yue
• Feature selection https://scikit-learn.org/stable/modules/feature_selection.html#feature-selection Jiahao
• Dimension reduction: PCA, ICA, CCA, FLD, t-SNE: https://www.datacamp.com/community/tutorials/introduction-t-sne Jiahao
• Cluster Jiahao

NLP-related Model Banghua

• Text
• Word2Vec

Common tools

• Numpy, Pandas, Pickle, Jupyter

Potential Collaboration choice:

• Github
• Google Drive

Prepare Overleaf Template

Potential machinary:

James-Stein estimation: adaptive estimator

Statsmodel gives an latex form automatically