RESERVE MY SPOT
Emeli Dral is a data scientist with large experience of using machine learning and data analysis techniques both in an industry and science. Emeli has a wide experience of leading machine learning projects for big companies in different domains vary from banking and telecommunications to e-commerce and hr-analytics. Emeli graduated from People’s friendship university of Russia, majoring in natural language processing. For more than 5 years she gives lectures about programming python, system engineering, machine learning and data analysis at different universities and summer schools. She is also a co-author of «Machine learning and data mining» specialization on Coursera.
• Identify cases where machine learning techniques should be applied
• Apply machine learning algorithms and techniques to real world applications
• Formulate problem statement and quality criteria
• Estimate potential economic effect of the machine learning models.
SKILLS:
- Machine Learning
- Mathematics
- MapReduce
- Statistics
- Data Analysis
- Bioinformatics
- Matlab
- Lecturing
- Python
- Text Mining
- Computer Science
- Data Science
- Natural Language Processing
- C++
DATE: 24–28 Apr, 2017
DURATION: 5 Days
LECTURES: 3 Hours per day
LANGUAGE: English
LOCATION: Barcelona, Harbour.Space Campus
COURSE TYPE: Offline
Session 1
Lifecycle of the typical machine learning project: from problem statement to quality evaluation.
WHAT YOU WILL LEARN
COURSE OUTLINEABOUT EMELIBIBLIOGRAPHYHARBOUR.SPACE Session 2
Recommender systems in industries:
product and media recommendations.
Session 3
Hybrid recommender systems: mixing, switching, cascading and other techniques.
Session 4
Time Series Forecasting, main components and attributes. Autocorrelation and stationarity. ARIMA models: structure, properties, fitting, evaluation. Automatic forecasting approaches for single and multiple time series.
INDUSTRIAL MACHINE LEARNING
The module covers topics related to industrial applications of machine learning. Nowadays machine learning technologies are widely used in practice in various applied fields such as retail, mass media, PR and marketing, banking, telecommunications, manufacturing, science and many other areas. Using relevant techniques in each project is very important, but often selecting a particular machine learning algorithm does not play a key role. Frequently the most important factors include relevant problem statement in terms of business goals, correct mathematical formalization of the problem, precise estimation of the potential economic effect, criteria and metrics of decision quality estimation and other factors.
In the course, we will learn the structure and the lifecycle of the machine learning project and cover topics ranging from the problem statement to final quality assessment as well as estimation of the economic effect.
EMELI DRAL
HARBOUR.SPACE UNIVERSITY
DATE: 24 –28 Apr, 2017
DURATION: 5 Days
LECTURES: 3 Hours per day
LANGUAGE: English
LOCATION: Barcelona, Harbour.Space Campus
COURSE TYPE: Offline
INDUSTRIAL MACHINE LEARNING
EVGENIY RIABENKO
VICTOR KANTOR
Session 5
Applied machine learning cases:
churn prediction and prevention, product recommender system, sentiment analysis, etc.
Evgeniy Riabenko is a data scientist with 10 years of experience in both industry and academia. He got his PhD in mathematical modelling from Moscow State University, and has expertise in statistics, machine learning, optimization, time series analysis, topic modelling, and bioinformatics. For 7 years he taught Statistics for Data Analysis course at Moscow State University, Moscow Institute of Physics and Technology and Higher School of Economics; an adapted version of the course is available on Coursera.
ABOUT EVGENIYVictor Kantor was born in 1992, graduated with honors at MIPT — Moscow Institute of Physics and Technology, the Department of Data Analysis (the basic organization – Yandex). Since 2011, Victor was engaged in data analysis and machine learning within various projects, and in 2012 he taught first as a teacher assistant, then held laboratory courses and subsequently lectured in data analysis courses.
In 2016, Victor co-developed the series of the most popular Data Analysis courses on Coursera attracting 30 thousand students around the world. Additionally Victor developed a course for "Data mining in action" at MIPT which became one of the most popular offline course on machine learning in Russia, attracting more than 600 participants. Victor worked as a data analysis specialist and as teacher for companies in different sectors (IT, retail, telecom) including Yandex, ABBYY, Sberbank Technologies, Megafon, 585gold, TSUM and others. Currently Victor works as researcher in Yandex and as a teaches at MIPT as well as an instructor and trainer for the master classes on machine learning within various companies.
ABOUT VICTOR
EVGENIY RIABENKO