ML engineer (ML-driven Java code optimization of Android apps)

компания "samsung research center" responsibilities: research and implement ml algorithms to process java bytecode research code features extraction decide on the ml solution architecture and training set research performance issues of ml algorithms implementation analysis of competitive ml solutions requirements: ba/bs in applied mathematics, applied statistics, computer science or similar 3 years of practical experience with common neural network / deep learning platforms (tensorflow, pytorch/caffe2) in a production environment understanding and hands-on experience with “classic” machine learning (k-nn, naïve bayes, svm, decision forests, etc.) and genetic ml algorithms experience with nlp technics (embedding, word2vec, bag-of-words, etc.) basic knowledge of c/c++ or java knowledge of data structures and algorithms complexity english (to write documentation and communicate by email) desired skills: knowledge and experience in structured prediction (structured svms, graph neural networks, etc.) experience with nn-based nlp (lstm, etc.) experience in sw development understanding of software design patterns understanding of compiler internals (e.g. llvm) and code optimizations, experience with code parsers deep understanding of computer system architecture scientific background (phd, patents, papers) work conditions: comfortable a-class office within 15 minute walking distance from savelovskaya and maryina roscha metro stations (shuttle bus available) flexible working time competitive salary full compliance with russian labor code corporate lunches medical and life insurance 7 paid sick leaves (without sick leave certificate) fitness room, shower, fitness compensation company compensates relocation and interview expenses samsung r&d institute do not intend to receive and use any information that constitutes or may constitute a trade (or other legally protected) secret of your former and/or current employers. you must not to disclose such information while applying and progressing through employment process.ооо «исследовательский центр самсунг» не заинтересовано в получении и использовании какой-либо информации, которая составляет или может составлять коммерческую (или иную охраняемую законом) тайну ваших бывших и/или текущего работодателей. вы не должны раскрывать такую информацию при отклике на вакансию и в процессе собеседования.