A Quick Start Getting to Know Me

A Quick Start Getting to Know Me

Tags
Computer Science
Career
Machine Learning
Author
ZHAO YUQI
A former Top Chinese Internet Company employee for several years and worked for smaller companies.
Passionate, Creative, Challenge-driven, Strongly aiming, Team coordinating, Bilingual communicating.
E-mailing for inquiry or further cooperation: [email protected]
 
In Brief:
A full-stack AI engineer in traditional and deep machine learning, serving widely in the internet field.
High-performing Algorithms & Machine Learning Engineer with 6 year+ experience in Anti-Cheat/Anti-Fraud Risk Controlling area, Internet Computational Advertising, Promotion coupons assignments, Growth Hacking, etc.
Fully aware of the whole pipeline of the Recommendation System; Unique understanding of the interaction of the Searching System; (those parts that are related to AI & machine learning)
Senior data scientist and big data developer with 9 year+ experience in internet finance/online e-commerce/O2O/blockchains.
 
Risk controlling(6 years+)
Anti-cheat risk includes fake advertising click/promotion cheaters/fictitious user identity, etc.
Anti-fraud is sometimes defined as protecting platform users from telephone fraud or other human-driven fraud on purpose.
Participate in building the entire pipeline of anti-cheat risk protection systems, including non-supervised pre-risk controlling/supervised discrimination during risk happening/risk judgment, to ensure that new risk types will be fed to the pre-stage immediately.
Significant experience in defining&data mining cheater behaviors.
Also able to provide a user rating system(FICO-like or bank) in financial risk control.
 
Growth Hacking(3 years+)
Provide abilities for both user quantity and quality rating systems from high dimensions, ensuring both efficiency and explanations based on user-level predictions, to help platforms to affect user preferences from an early stage.
Considering risk controlling from a growth point of view to ensure the user's lifelong time value(LTV or ROI).
 
Internet Computational Advertising(2 year+)
Deeply understanding the formula of Computational Advertising, including real-time API (RTA)/real-time bidding(RTB)/oCPX/DSP eCPM competition/user preference portfolio, participating in building a whole system as an advertiser with advertisement cost over $500k/per day. (those parts that are related to AI & machine learning)
 
Promotion coupons assignments(3 year+)
Experienced in assigning diverse promotions/coupons in several scenes, helping save financial budget, with full awareness of deriving math equations for the specific improving target, estimating prediction, and programming tools in math.
 
Recommendation & Searching System
Not directly participating but fully aware of Taobao's Recommendation System, deep academic understanding in multi-source recalling/collaborative filtering(U2U & I2I)/sequential ranking/re-ranking. Expert in multi-domain/multi-modal & ESMM deep learning model structures.
Re-enforcement learning point of view of defining the long interaction of users' long-term searching behaviors.
 
Technical directions that have landed in real scenes:
Supervised Learning, Multi-domain learning: Transfer Learning/Meta Learning/Incremental Learning/Multi-task Learning
Supervised Learning, Graph learning: Graph ML, e.g., GCN
Semi-supervised Learning: (Graph-based)Community Search, (Sequence-based)Page-Rank
Non-supervised Learning: (Graph-based)Community Detection, (Point-based)Anomaly Detection,(Time-based)Time series Forecasting, and Clustering for cold start situations.
Programming: Linear and Integer programming on non-convex datasets.
 
Big Data development skills:
Hive SQL/Spark with pyspark/Max Compute developing.