專家介紹
德國理論物理博士
現職於德國的數據科學家，四年工作經驗
技能:
• Python packages: numpy, sklearn, scipy, statsmodels, pandas, tensorflow.
• Computer vision:
o Object detection: Mask-RCNN, YOLO (AlexeyAB darknet).
o Image enhancement: Image super resolution (ISR), perspective sharpness index.
• Natural language processing: BERT (simple transformer), word2vec, topic modeling (gensim)
• Extra machine learning packages:
o Supervised learning: xgboost, catboost, lightgbm
o Unsupervised: matrix factorization (funk-svd, lightfm),
o Hyperparameter tuning: BayesianOptimization
• Web crawling: selenium, bs4.
• Time series analysis: SARIMA, prophet, seq2seq (encoder-decoder LSTM).
