// PROJECT DETAIL
书价预测
已完成2025-06 — 2025-06Data Scientist / ML Engineer
Predict book prices from metadata like ratings, genres, and descriptions using ML.
核心功能
- ▸Data loading and merging
- ▸Preprocessing including binarization, one-hot encoding, imputation, NLP cleanup
- ▸Feature engineering: age from publish date, star rates from ratingsByStars
- ▸Feature selection via correlation and log-transform
- ▸Model training with Neural Network and LightGBM
技术亮点
- ▸Swapped train/test naming handled
- ▸End-to-end pipeline with two models: Neural Network and LightGBM
- ▸Text feature extraction using TF-IDF and TruncatedSVD
技术栈
PythonJupyterpandasnumpyscikit-learnNLTKTensorFlow/KerasLightGBM
关联文章
暂无关联文章
日历事件
// Phase 4 接入