vOtpuskSam.ru
vOtpuskSam.ru
AI powered daily feed of YouTube travel videos on a map, based on Natural Language Processing and Named Entity Recognition.
NLP, NER, Python, Google Maps API, JS
vOtpuskSam.ru
Read more
Apache Superset setup for Edge Capital Management
Apache Superset setup for Edge Capital Management
2021
Apache Superset BI-system installation for a macro DeFi-hedge fund (Edge Capital Management LLC) for data explorations and researches in the web3.
Business Intellegence System.
Project of 2021.
Read more
MIPT courses
Models for various DS/ML courses
2020
Building DS/ML models for the clients to complete various courses (mostly from the MIPT - Moscow Institute of Physics and Technology).
Developed more than 10 custom models: used car prices predictions, detection of false tweets with TF-IDF, W2v and NN (Pytorch), image classification with CNN, tweet sentiment analysis, etc.
Read more
Proppant granuls
Computer Vision Proppant Check Challenge
2020
A Rosneft hackathon to predict number of proppant granuls on a picture. Organized by Boosters.pro.
Tools:
Pytorch, CNN, Python
Read more
Two sigma
TWO SIGMA: MY SILVER MEDAL SOLUTION OF THE KAGGLE CONTEST “USING NEWS TO PREDICT STOCK MOVEMENTS”
2019
My silver medal solution of a Kaggle challenge from Two Sigma, 2019. The aim of the contest was to do the best stock movement prediction based on news and market data.
NLP
Read more
Kaggle
Kaggle competitor
Participation in Kaggle competitions (2012, 2018 -2020).
- Status: "Notebooks expert"
1 silver medal for a championship, 6 medals for notebooks. - Tools:
Python (Pandas, Numpy, Scikit-learn etc.), Pytorch, Tensorflow/Keras, LightGBM, XGBoost, Logit, PCA, k- means etc.
Read more
Data Science Lead at MegaFon
Data Science Lead at the top telecom company (MegaFon)
I has a 6 years experience as a Data science lead at a top 3 telecom companies in Russia (MegaFon).
2008 - 2014
Projects accompished by me:
- Leading a big team of data scientists.
- Organized all the CRM-campaigns in the company (100 launches a month).
- Built models for: churn-prediction, cross-sell, credit scoring, gender prediction etc.
- Performed a lot of clustering models for customers segmentation.
Read more
Commercial projetcs
Examples of data science models built
- Credit scoring in telecom
- Churn prediction in telecom
- Clustering models for customers segmentation.
- Predicting the gender of telecom-subscribers
- Relationship between the macro indicies and macroeconomic factors
- Forecasting revenue and sales in telecom
Read more