- Level Professional
- Duration 8 hours
- Course by SAS
-
Offered by
About
This course focuses on data exploration, feature creation, and feature selection for time sequences. The topics discussed include binning, smoothing, transformations, and data set operations for time series, spectral analysis, singular spectrum analysis, distance measures, and motif analysis. In this course you learn to perform motif analysis and implement analyses in the spectral or frequency domain. You also discover how distance measures work, implement applications, explore signal components, and create time series features. This course is appropriate for analysts with a quantitative background as well as domain experts who would like to augment their time-series tool box. Before taking this course, you should be comfortable with basic statistical concepts. You can gain this experience by completing the Statistics with SAS course. Familiarity with matrices and principal component analysis are also helpful but not required.Modules
Welcome
1
Videos
- Overview
1
Readings
- Getting the Most from this Specialization
Introduction to Time Series Mining and Creation
1
External Tool
- Access the Virtual Lab
1
Videos
- Meet the Instructor
2
Readings
- Prerequisites
- Finding the Course Files and Practicing in this Course (UPDATED for Virtual Lab)
Time Series Essentials
2
Videos
- Introduction to Time Series
- Time Series Data Creation
Accumulation, Exploration and Binning
1
Assignment
- Think About It - Handling Missing Values
7
Videos
- Selecting an Interval for Accumulation, Part 1
- Selecting an Interval for Accumulation, Part 2
- Demo: Accumulating Transactional Data to Time Series
- Summary Measures on Time Series
- Demo: Exploring Time Series Summary Characteristics
- Binning Time Series
- Demo: Exploring Time Series Using Binning
Signal
4
Videos
- Signal versus Noise
- Signal Types and Decompositions
- Demo: Decompositions Using the TSA Package
- Demo: Feature Creation Using the DATA Step-like Syntax in the TSMODEL Procedure
Practice What You Learned
3
Assignment
- Practice: Explore and Accumulate a Time Series
- Practice: Create Smoothed Representations of Time Series
- Practice: Perform a Decomposition Analysis
Sequence Distance Basics
2
Assignment
- Question - Evaluating Paths
- Question - Warping
1
External Tool
- Access the Virtual Lab
5
Videos
- Introduction to Distance Measures for Time Series
- Measuring Sequence Similarity
- Direct Mapping and Time Warping
- Demo: An Example of Time Warping and Relative Path Costs
- Demo: Time Series Clustering
Symbolic Representation of Sequences
1
Assignment
- Think About It - SAX Word Sequences
3
Videos
- Introduction to Symbolic Representation of Sequences
- The Symbolic Aggregate ApproXimation (SAX)
- Demo: SAX Distance for Input Variable Ranking Based on Similarity to the Target
Practice What you Learned
2
Assignment
- Practice: Perform a Similarity Analysis, Part 1
- Practice: Perform a Similarity Analysis, Part 2
Spectral Analysis Basics
1
Assignment
- Question - Spectral Analysis
1
External Tool
- Access the Virtual Lab
3
Videos
- Introduction to Spectral Analysis
- Spectral Analysis Fundamentals
- Demo: Introduction to Spectral Analysis
The Periodogram and Spectral Density
2
Assignment
- Question - Sine Waves
- Think About It - Periodogram
4
Videos
- Basic Ideas from Trigonometry
- A Regression Approach to Cycle Identification
- Estimating the Spectral Density
- Demo: Implementing Information from the Periodogram
Singular Spectrum Analysis
1
Assignment
- Think About It - Window Size in SSA
4
Videos
- An Introduction to Singular Spectrum Analysis (SSA)
- SSA Step-by-Step
- Demo: Singular Spectrum Analysis
- Demo: Application of Univariate Singular Spectrum Analysis
Practice What You Learned
2
Assignment
- Practice: Perform and Interpret a Fourier Analysis
- Practice: Perform and Interpret a Singular Spectrum Analysis
Motif Analysis Basics
1
Assignment
- Think About It - Applications of Motif Analysis
1
External Tool
- Access the Virtual Lab
6
Videos
- Introduction to Motif Analysis
- Motif Discovery Basics
- Two Approaches to Motif Discovery
- Demo: Motif Discovery Using the Brute Force Method
- Demo: Motif Discovery Using the Probabilistic Model Method
- Demo: Motif Scoring
Practice What You Learned
1
Assignment
- Practice: Score a Table
Review Quiz
1
Assignment
- Creating Features for Time Series Data - Course Exam
Auto Summary
"Creating Features for Time Series Data" is a professional-level course offered by Coursera, focusing on data exploration, feature creation, and selection for time series analysis. Ideal for analysts and domain experts with a quantitative background, it covers binning, smoothing, transformations, spectral and singular spectrum analysis, distance measures, and motif analysis. The course spans 480 minutes and requires familiarity with basic statistical concepts, with a recommendation to complete the Statistics with SAS course beforehand. Subscription is available through the Starter plan.
Chip Wells