Lead data scientist

Role : Senior Data Scientist / Lead Data ScientistAbout Media.net : Media.net is a leading, global ad tech company that focuses on creating the most transparent and efficient path for advertiser budgets to become publisher revenue. Our proprietary contextual technology is at the forefront of enhancing Programmatic buying, the latest industry-standard in ad buying for digital platforms.The Media.net platform powers major global publishers and ad-tech businesses at scale across ad formats like display, video, mobile, native, as well as search. Media.net’s U. S. HQ is based in New York, and the Global HQ is in Dubai. With office locations and consultant partners across the world, Media.net takes pride in the value-add it offers to its 50+ demand and 21 K+ publisher partners, in terms of both products and services.Data Science is at the heart of Media.net. The team uses advanced statistical and machine learning and deep learning models, large scale distributed computing along with tools from mathematics, economics, auction theory to build solutions that enable us to match users with relevant ads in the most optimal way thereby maximizing revenue for our customers and for Media.net.Some of the challenges the team deals with : How do you use information retrieval, machine learning models to estimate click through rate and revenue given the information regarding the position of the slot, user device, location and content of the page. How do you scale the same for thousands of domains, millions of urls?How do you match ads to page views considering contextual information? How do you design learning mechanisms to continuously learn from user feedback in the form of clicks and conversions? How do you deal with the extremely sparse data? What do you do for new ads and new pages? How do we design better explore-exploit frameworks? How do you design learning algorithms that are fast and scalable?How do you combine contextual targeting with behavioral user-based targeting?How do you establish a unique user identity based on multiple noisy signals so that behavioral targeting is accurate?Can you use NLP to find more genetic trends based on the content of the page and as?What is in it for you?Understand business requirements, analyze and extract relevant information from large amounts of historical data.Use your knowledge of Information retrieval, NLP, Machine Learning (including Deep Learning) to build prototype solutions keeping scale, speed and accuracy in mind.Work with engineering teams to implement the prototype. Work with engineers to design appropriate model performance metrics and create reports to track the same.Work with the engineering teams to identify areas of improvement, jointly develop research agenda and execute on the same using cutting edge algorithms and tools.You will need to understand a broad range of ML algorithms and appreciation on how to apply them to complex practical problems. You will also need to have enough theoretical background and a good grasp of algorithms to be able to critically evaluate existing ML algorithms and be creative when there is a need to go beyond textbook solutions.Who should apply for this role ?Ph D / Research Degree or BS / MS in Computer Science, Statistics, Artificial Intelligence, Machine learning, Operations Research or related field.6 - 8 years of experience in building Machine Learning / AI / Information Retrieval modelsExtensive knowledge and practical experience in machine learning, data mining, artificial intelligence, statistics.Understanding of supervised and unsupervised algorithms including but not limited to linear models, decision trees, random forests, gradient boosting machines etc.Excellent analytical and problem-solving abilities.Good knowledge of scientific programming in Python.Experience with Apache Spark is desired.Excellent verbal & written communication skillsBonus Points : Publications or presentation in recognized Machine Learning and Data Mining journals / conferences such as ICMLKnowledge in several of the following : Math / math modeling, decision theory, fuzzy logic, Bayesian techniques, optimization techniques, statistical analysis of data, information retrieval, natural language processing, large scale data processing and data miningAbility deal with ambiguity & break them down into research problemsStrong theoretical and research acumen