Tech Stack: Scikit-learn, XGboost, Catboost, NLTK, Transformers, Torchtext, Fasttext, Gensim, SpaCy, OpenCV, Keras, Tensorflow, Pytorch, Torchvision, Pillow, Eviews, Jdemetra, Prophet, GluontTS, ETNA, Darts

Expert’s cases:

1. Rewriting Matlab trading simulation on Python

   • Speed up of the Simulation ~1.8 times.

   • Implementation of Matlab behavior during training/evaluation, on Python (Metrics, inputs, outpus, etc.)

2. UI creation for BI dashboards

   • Data validation

   • Notebook creation for two special structure comparison

3. Performed Machine Translation from different languages to English

   • Performed Resume Classification by department

   • Implementation of Resume NER

   • Created reports on training, validation; statistics before training

4. Performed Sentiment Analysis on people’s reviews and matching the results with their rates

   • EDA of existing dataset

   • Performed Data Labeling (We don’t know for sure whether the reviews positive or negative)

   • Сlustering of vector representations of reviews