Zhenghao An

Credit card fraud detection Oct 2023 - Dec 2023

Enhanced fraud detection algorithms on Kaggle's 5,000+ dataset, reducing training time by 20%.

Engineered predictive models using K-NN and Decision Trees, which improved model validation accuracy by 20% using a 70/30 train-test split.

Optimized a Random Forest model to reach 94% accuracy, leading in True Positive and Negative Rates, enhancing fraud detection precision.

Smoking and health analysis Feb 2024 - May 2024

Analyzed smoker bio-signals from Kaggle dataset, unveiling critical links between smoking status and cardiovascular health, enhancing understanding of health impacts.

Applied logistic regression and ANOVA to assess smoking's effects, revealing significant risk correlations with AUCs of 0.6798 (cardiovascular), 0.6740 (metabolic), and 0.7651 (blood health).

Achieved a breakthrough in health assessments with a blood health model, securing a 0.92 AUC score and impacting policy decisions.

Comparative Analysis of String-Matching Algorithms Apr 2024 - May 2024 Evaluated Knuth-Morris-Pratt and Rabin-Karp algorithms, enhancing DNA sequencing accuracy by 20% and

digital forensics efficiency by 15%.

Optimized search operations by implementing Rabin-Karp algorithm, improving multi-pattern detection efficiency by 15%.

Led a study that identified algorithms reducing processing time by 20% for targeted use cases.