Biography

Valentina Djordjevic works as a Data Scientist at Things Solver. She has a Bachelor’s degree in Information Systems and Technologies and a Master’s degree in Business intelligence, at University of Belgrade, Faculty of Organizational Sciences. The main fields of studies she focuses on the most are time series analysis and anomaly detection techniques. She has a strong technical knowledge in the field of Data Science, including programming (Java, Python, SQL, ETL), statistics (descriptive statistics, hypothesis testing, probability theory,…), modeling (machine learning algorithms- neural networks, random forest, linear regression, k-means, isolation forest, association rules, recommender systems, ARIMA models…) and visualization (Matplotlib, Plotly, Tableau,…). Data science problems that she’s been working on are coming from various business domains, from telecommunications to retail and banking, where she’s dealing with forecasting, predictive maintenance, anomaly detection, segmentation, churn prevention, etc.

Workshop

Time Series Forecast

This workshop will cover basic concepts of time series analysis, like time series decomposition, stationarity analysis, trend and seasonality smoothing. Afterwards, some of the most popular algorithms used for time series forecasting will be presented and explored. The workshop will include programming in Python, and its time series forecasting library – PyFlux.

Agenda

• Intro to time series analysis:
  – What is a time series
  – Time series decomposition
  – Stationarity testing
• Intro to Pyflux library
  – Basic concepts and general overview of its functionalities
  – Exploring models and their characteristics, assumptions and parameters
• Hands-on example
  Create several forecasting models (ARIMA, VAR, GARCH,…)
  • divide a set into train and test sets
  • define a model
  • fit a model
  • extract model results and estimate model perfomance
  • comparative analysis of different models.

Course objectives

The main goal of this workshop is to introduce participants with main concepts of time series analysis, as well as with forecasting methods available in the PyFlux library.

Target audience

The target audience includes those interested in time series forecasting techniques, examples and applications. Everybody who loves programming as well as data science is invited to come, and learn something new.

Course prerequisites

– Installations:

            – Anaconda environment with Python 2.7, and Pyflux library installed on it.

– Technical knowledge:

            – Basic programming in Python, or some other programming language

            – Basic math knowledge.