This course will guide you step by step through many challenges that modern data scientists face in modern world while choosing right model, choosing right metrics and answering important business questions. You will learn how to assess real NLP tasks from scratch. You will create and train linear, tree-based and deep learning models, compare their strengths and weaknesses. You will learn abstractions of top-notch deep learning architectures on the edge of today science.
- Data Preprocessing: main text preprocessing techniques, feature engineering, and word embeddings
- Linear models and tree ensembles for sentiment analysis task
- Recurrent Neural Network and its modifications for text classification
- Word- and char- based convolution neural networks
- Advanced architectures, tricks, and pitfalls
Python 3, NN framework: keras
Basic linear algebra, proficiency in Python, machine learning basic (linear- and tree-based models), a basic understanding of neural networks.
Data enthusiast at Ciklum/ Ukraine Data Science Club
Former actuary at Metlife. Data Scientist in Leboutique, 2know. Co-creator @ datafest.in.ua.
Fields of interests: Computer Vision, NLP, Deep Learning
Data Science Intern at Ciklum, Master Student at UCU
I am fond of Data Science for 2 years and can say that it is my hobby. During this time I worked in retail, internet advertisement and outsourcing company Ciklum. Now I’m studying at Data Science Master program at UCU. I was a winner of several Data Science competitions. In 2016 I held Kyiv Kaggle Training and UCUracy Kaggle training. In 2017 with my college Alex Ruppelt we created Ukrainian Data Science Club and organized DataFest Kyiv 2017. Moreover, I was a mentor in Yandex and MIPT “Machine learning and data analysis.” courses on Coursera, article author in ODS open ml course.
Fields of interests: NLP, Deep Learning, Time Series, Xgboost 🙂