What will the future hold?
From inventory to website visitors, resource planning to financial data, time-series data is all around us. But how can you know what the future holds? This two-day course empowers you to go beyond “spotting trends” and make data-driven business forecasts.
What you'll learn
- How to effectively handle time-series data
- Python utilities that make working with time-series a breeze
- Why model validation with time-series data cannot follow the traditional machine learning methodology
- How to determine which model best suits particular time series data
- Why feature engineering is fundamental to the success of your modeling
- How to incorporate seasonality into your models
Program
The program consists of eleven blocks. Each block consists of a theory component and a hands-on lab.
Day 1:
- Time features encoding and formatting;
- Pandas time series features (smoothing, resampling, re-weighting);
- Sessionization and holiday detection
- Feature Engineering for time series
- Additive vs Multiplicative features
- Error-Trend-Seasonality Decomposition;
Day 2:
- Seasonality estimation;
- Forecast evaluation and model selection;
- Forecasting with Prophet;
- Switch-point Detection;
- Outlier Detection.
Climbing a steep Python and Machine Learning curve in three days. This would have taken me months on my own.
This course is perfect for
Data Scientists who know Machine Learning and want to expand their skillset by moving from static data sets to dynamic time-dependent data sets.
We invite anyone who is familiar with Python and statistics and wants to become more productive and empowered in analyzing and forecasting using time-series data to join us. To ensure you get the most out of these two days, we recommend you have at least one year of work experience with pandas, scikit-learn, and Matplotlib.
What will you learn during Practical Time Series Analysis & Forecasting?
You will learn to confidently work with time-series data: cleaning it, removing outliers, and handling missing data. You will also learn how to create forecasts with your data sets and validate your models when using time-series data.
Data Science Skills
The Data Science Learney Journey
The Right Format For Your Preferred Learning Style
At GoDataDriven we offer four distinct training modalities:
- In-Classroom & In-Company Training
- Online, Instructor-Led Training
- Hybrid and Blended Learning
- Self-Paced Training
