Boris Sharchilev

FAQ About Boris Sharchilev

• What is Boris Sharchilev's address?

  Boris Sharchilev is located at
  Moscow, Moscow City, Russian Federation

• What is Boris Sharchilev's job title?

  Boris Sharchilev is a
  Head of Zen Publishing platform Engineering Department. Leading 4 engineering teams (~30 engineers overall).Leading a team of 8 engineers responsible for: - Traffic management @ Zen publishing platform - Ad recommendations. Working on recommendations' quality. Highlights: - New algorithm for exploration of fresh publications: massive user metrics increase - New "author-centric" algorithm satisfying business goals without decreasing product metrics * Paper "Web-Based Startup Success Prediction" accepted at CIKM 2018. Highlights: - Finding influential samples for Neural Networks: increased intent recall @Yandex.Alice, implemented in production - Finding influential samples for Gradient Boosted Decision Trees: * Paper "Finding Influential Training Samples for Gradient Boosted Decision Trees" accepted at ICML 2018 * Method implemented as a separate mode in CatBoost (https://tech.yandex.com/catboost/doc/dg/concepts/cli-reference_ostr-calc-docpage/). Performing Machine Learning research applied to Information Retrieval and Web Search problems as a part of Yandex.Research team. PhD student in a joint programme by Yandex and University of Amsterdam. Highlights: - Promising startup recommendation: SOTA results compared to known literature - Interpreting Gradient Boosted Decision Trees ensemble: large metrics boost @Yandex.Zen. Conducted lab and tutorial problem - solving session for Master - level students in modules related to probability theory and statistics, data analysis, estimation and decision theory: - ACS424 Multisensor Data Fusion - http://www.sheffield.ac.uk/acse/current/modules/acs424 - ACS6124 Multisensor and Decision Systems - http://www.sheffield.ac.uk/acse/current/modules/acs6124. Big data and Image Processing activity - Implemented parallel image filtering/smoothing algorithms using Apache Hadoop. - Implemented and designed a fast 3D - 2D very large image registration algorithm (Up to 100GB images). - Implemented and designed a custom Graph-Cut-based uCT image segmentation algorithm. - Performed morphological analysis of rock uCT images, introduced and implemented new connectivity/permeability morphological metrics.Was involved in development of web and user interfaces and in building architecture of a custom task manager in Objective-C, performed basic SQL database management tasks. Facebook

• What industry does Boris Sharchilev currently work in?

  Boris Sharchilev currently works in the
  Facebook

• What is Boris Sharchilev's area of expertise?

  Boris Sharchilev specializes in
  Head of Zen Publishing platform Engineering Department. Leading 4 engineering teams (~30 engineers overall).Leading a team of 8 engineers responsible for: - Traffic management @ Zen publishing platform - Ad recommendations. Working on recommendations' quality. Highlights: - New algorithm for exploration of fresh publications: massive user metrics increase - New "author-centric" algorithm satisfying business goals without decreasing product metrics * Paper "Web-Based Startup Success Prediction" accepted at CIKM 2018. Highlights: - Finding influential samples for Neural Networks: increased intent recall @Yandex.Alice, implemented in production - Finding influential samples for Gradient Boosted Decision Trees: * Paper "Finding Influential Training Samples for Gradient Boosted Decision Trees" accepted at ICML 2018 * Method implemented as a separate mode in CatBoost (https://tech.yandex.com/catboost/doc/dg/concepts/cli-reference_ostr-calc-docpage/). Performing Machine Learning research applied to Information Retrieval and Web Search problems as a part of Yandex.Research team. PhD student in a joint programme by Yandex and University of Amsterdam. Highlights: - Promising startup recommendation: SOTA results compared to known literature - Interpreting Gradient Boosted Decision Trees ensemble: large metrics boost @Yandex.Zen. Conducted lab and tutorial problem - solving session for Master - level students in modules related to probability theory and statistics, data analysis, estimation and decision theory: - ACS424 Multisensor Data Fusion - http://www.sheffield.ac.uk/acse/current/modules/acs424 - ACS6124 Multisensor and Decision Systems - http://www.sheffield.ac.uk/acse/current/modules/acs6124. Big data and Image Processing activity - Implemented parallel image filtering/smoothing algorithms using Apache Hadoop. - Implemented and designed a fast 3D - 2D very large image registration algorithm (Up to 100GB images). - Implemented and designed a custom Graph-Cut-based uCT image segmentation algorithm. - Performed morphological analysis of rock uCT images, introduced and implemented new connectivity/permeability morphological metrics.Was involved in development of web and user interfaces and in building architecture of a custom task manager in Objective-C, performed basic SQL database management tasks.

• Which languages does Boris Sharchilev speak? 

  Boris Sharchilev speaks the following languages:
  No languages available.

• Which industries has Boris Sharchilev previously worked in?   

  Boris Sharchilev has prior experience in the following industries:
  Yandex , Yandex , Yandex , Yandex , Yandex , Yandex , University of Sheffield , Schlumberger , AT Consulting

• What is Boris Sharchilev’s work experience?  

  Boris Sharchilev has the following professional experience:
  Facebook , Yandex , Yandex , Yandex , Yandex , Yandex , Yandex , University of Sheffield , Schlumberger , AT Consulting .

• How can I find Boris Sharchilev on LinkedIn?  

  You can view Boris Sharchilev’s
  LinkedIn profile at LinkedIn URL.

• What are Boris Sharchilev’s work email addresses?  

  Boris Sharchilev’s work email addresses are
  [email protected] .

• What company does Boris Sharchilev work for?  

  Boris Sharchilev works at
  Facebook