Modeling Time Series and Sequential Data
In this course you learn to build, refine, extrapolate, and, in some cases, interpret models designed for a single, sequential series. There are three modeling approaches presented. The traditional, Box-Jenkins approach for modeling time series is covered in the first part of the course. This presentation moves students from models for stationary data, or ARMA, to models for trend and seasonality, ARIMA, and concludes with information about specifying transfer function components in an ARIMAX, or time series regression, model. A Bayesian approach to modeling time series is considered next.
Monthly Subscription Starting at AED 99 + VAT
- Level Professional
- المدة 11 ساعات hours
- الطبع بواسطة SAS
-
Offered by
عن
In this course you learn to build, refine, extrapolate, and, in some cases, interpret models designed for a single, sequential series. There are three modeling approaches presented. The traditional, Box-Jenkins approach for modeling time series is covered in the first part of the course. This presentation moves students from models for stationary data, or ARMA, to models for trend and seasonality, ARIMA, and concludes with information about specifying transfer function components in an ARIMAX, or time series regression, model. A Bayesian approach to modeling time series is considered next. The basic Bayesian framework is extended to accommodate autoregressive variation in the data as well as dynamic input variable effects. Machine learning algorithms for time series is the third approach. Gradient boosting and recurrent neural network algorithms are particularly well suited for accommodating nonlinear relationships in the data. Examples are provided to build intuition on the effective use of these algorithms. The course concludes by considering how forecasting precision can be improved by combining the strengths of the different approaches. The final lesson includes demonstrations on creating combined (or ensemble) and hybrid model forecasts. This course is appropriate for analysts interested in augmenting their machine learning skills with analysis tools that are appropriate for assaying, modifying, modeling, forecasting, and managing data that consist of variables that are collected over time. This course uses a variety of different software tools. Familiarity with Base SAS, SAS/ETS, SAS/STAT, and SAS Visual Forecasting, as well as open-source tools for sequential data handling and modeling, is helpful but not required. The lessons on Bayesian analysis and machine learning models assume some prior knowledge of these topics. One way that students can acquire this background is by completing these SAS Education courses: Bayesian Analyses Using SAS and Machine Learning Using SAS Viya.الوحدات
Welcome
-
1
Videos
-
1
Readings
1
Videos
- Overview
1
Readings
- Getting the Most from this Specialization
Course Overview and Logistics
-
1
Videos
-
2
Readings
-
1
ExternalTool
1
External Tool
- Access the Virtual Lab
1
Videos
- Welcome to the course
2
Readings
- Prerequisites
- Finding the Course Files and Practicing in this Course (UPDATED for Virtual Lab)
Module Overview
-
1
Videos
1
Videos
- About this Module
A Review of Time Series Components and Concepts
-
5
Videos
-
1
Assignment
1
Assignment
- Question: Statistical Time Series
5
Videos
- Time Series Components
- Applications of Time Series Analysis
- Demo: Exploring a Time Series
- A Framework for Forecasting
- Demo: Accumulating a Time Series and Exploring Systematic Variation
Simple Models for Time Series
-
2
Videos
2
Videos
- Concepts and Notation
- Naive Models
Exponential Smoothing Models
-
3
Videos
-
1
Assignment
1
Assignment
- Practice: Forecasting with ESMs
3
Videos
- Introduction to Exponential Smoothing Models (ESM)
- ESM and Signal Components
- Demo: Forecasting with ESM
Module Overview
-
1
Videos
-
1
ExternalTool
1
External Tool
- Access the Virtual Lab
1
Videos
- About this Module
ARMA Models
-
7
Videos
-
1
Assignment
1
Assignment
- Question: Stationary Time Series
7
Videos
- Models for Stationary Data
- Autoregressive Moving Average Models
- Identifying ARMA Models (Part 1)
- Identifying ARMA Models (Part 2)
- Demo: ARMA Model Properties
- Automatic Order Identification
- Demo: Identifying ARMA Orders
ARIMA Models, Trend
-
7
Videos
7
Videos
- Non-Stationary Data, Trend
- Differencing and Integration
- Trend Functions
- Demo: Trend Two Ways in an ARIMA Framework
- The Augmented Dickey Fuller Unit Root (ADF) Test (Part 1)
- The Augmented Dickey Fuller Unit Root (ADF) Test (Part 2)
- Demo: An Application of the ADF Test
ARIMA Models, Seasonality
-
4
Videos
4
Videos
- Seasonal Variation (Part 1)
- Seasonal Variation (Part 2)
- The ADF Test for Seasonality
- Demo: Seasonality Two Ways in an ARIMA Framework
ARIMAX Models, Inputs
-
7
Videos
-
1
Assignment
1
Assignment
- Practice: ARIMAX - Identification, Estimation and Forecasting
7
Videos
- Time Series Regression
- Demo: Ordinary Regression Using Outliers
- The Cross Correlation Function (CCF)
- The Transfer Function
- Interpreting the CCF
- Demo: Dynamic Regression with Event Variables
- Cross Correlation Pitfalls
Module Overview
-
2
Videos
-
1
ExternalTool
1
External Tool
- Access the Virtual Lab
2
Videos
- About this Module
- Classical Analysis versus Bayesian Analysis
Bayesian Time Series Structure
-
4
Videos
-
4
Assignment
4
Assignment
- Question: PROC MCMC Diagnostics
- Question: PROC MCMC Statements
- Practice: Modeling Autoregressive Components in Concert Data
- Practice: Modeling Seasonality Components in Stock Data
4
Videos
- Accessing Lag and Next Values
- Demo: Setting Up Autoregressive Components
- Dynamic Linear Model Setup
- Demo: Setting Up Seasonality Components
Exogenous Variables
-
2
Videos
-
2
Assignment
2
Assignment
- Question: PROC MCMC Syntax
- Practice: Modeling Exogenous Components in Rose Sales Data
2
Videos
- Adding Exogenous Variables
- Demo: Setting Up Exogenous Components
Forecasting in Bayesian
-
2
Videos
-
2
Assignment
2
Assignment
- Question: Bayesian Scoring/Forecasting Techniques
- Practice: Generating Posterior Predictive Distributions for an AR(1) Model
2
Videos
- PREDDIST and Forecasting
- Demo: Forecast Output
Module Overview
-
1
Videos
-
1
ExternalTool
1
External Tool
- Access the Virtual Lab
1
Videos
- About this Module
Using Machine Learning Models for Time Series Forecasting
-
3
Videos
-
1
Assignment
1
Assignment
- Question: Machine Learning Models
3
Videos
- Preparing Time Series Data for Machine Learning
- Brief Introduction to Gradient Boosting Models
- Demo: Preparing Time Series Data and Building a Gradient Boosting Model"
Deep Learning with Recurrent Neural Networks for Time Series Forecasting
-
4
Videos
-
1
Readings
-
4
Assignment
4
Assignment
- Question: Data for Recurrent Neural Network Modeling
- Practice: Changing the Number of Lagged Input Values for the Recurrent Neural Network Model
- Practice: Adding Hidden Units to the Recurrent Neural Network Model
- Practice: Removing a Hidden Layer from the Recurrent Neural Network Model
4
Videos
- Introduction to Recurrent Neural Networks
- Long Short-Term Memory Blocks in RNNs
- Demo: Building a Recurrent Neural Network with LSTM Blocks to Forecast Time Series
- Limitations of Machine Learning Methods for Time Series Forecasting
1
Readings
- About the Next Three Practices
Module Overview
-
1
Videos
-
1
ExternalTool
1
External Tool
- Access the Virtual Lab
1
Videos
- About this Module
A Hybrid or Ensemble Approach to Forecasting
-
5
Videos
-
1
Assignment
1
Assignment
- Practice: Generating a Combined Model Forecast
5
Videos
- External Models and Combined Forecasts
- Combination Forecast Details
- Combined Forecasts Using the TSM Package
- Demo: Generating Combined Forecasts with the CFC Object
- Demo: Combining Forecasts from Multiple Modeling Approaches
Combining Traditional Time Series Methods with Machine Learning Methods
-
3
Videos
3
Videos
- Strengths of Machine Learning Methods: Modeling Multiple Time Series
- Weighting Combined Forecasts with Machine Learning
- Demo: Using Gradient Boosting to Find the Best Weighted Combination of Traditional Time Series Models
Review
-
1
Assignment
1
Assignment
- Modeling Time Series and Sequential Data - Course Exam
Auto Summary
Unlock the secrets of time series and sequential data modeling with the "Modeling Time Series and Sequential Data" course. Designed for data science and AI enthusiasts, this professional-level course, available on Coursera, offers a comprehensive exploration of advanced modeling techniques. Guided by expert instructors, you will delve into three primary approaches for time series analysis. Begin with the traditional Box-Jenkins methodology, mastering ARMA and ARIMA models, and learning to incorporate transfer function components in ARIMAX models. Then, transition to Bayesian models, extending the basic framework to handle autoregressive variations and dynamic input variables. Finally, harness the power of machine learning algorithms, including gradient boosting and recurrent neural networks, to manage nonlinear data relationships effectively. The course emphasizes practical application, using software tools such as Base SAS, SAS/ETS, SAS/STAT, and SAS Visual Forecasting, alongside open-source tools. While prior knowledge of Bayesian analysis and machine learning is beneficial, it is not mandatory, making the course accessible to a wide range of learners. Spanning 660 minutes of in-depth content, the course culminates in advanced strategies for enhancing forecasting precision through ensemble and hybrid models. Ideal for analysts aiming to elevate their machine learning capabilities, this course offers both Starter and Professional subscription options to suit varied learning needs. Whether you're looking to refine your existing skills or expand your analytical toolkit, "Modeling Time Series and Sequential Data" provides the knowledge and resources necessary to master this critical area of data science. Join today and take a significant step towards becoming an expert in time series modeling and forecasting.
Chip Wells
Ari Zitin
Danny Modlin