NLP for Financial Applications
Investigate DL/RL methods for processing overlength, noisy and obscure financial-related text, e.g. financial news, reports, posts, etc., particularly on sentiment analysis, summarization, NER and etc.
Financial Behavioral Data Mining
Investigate supervised, semi-supervised and unsupervised learning methods for temporal, heterogenous and large-scale financial behavior data, especially for anomaly detection applications.
Refinement for Online Ads
Investigate machine learning methods for improving CTR/CVR of online ads recommendations, with special focus on cold-start prediction, model calibrations and delayed feedback modeling.