Book Image

Modern Time Series Forecasting with Python

By : Manu Joseph
5 (1)
Book Image

Modern Time Series Forecasting with Python

5 (1)
By: Manu Joseph

Overview of this book

We live in a serendipitous era where the explosion in the quantum of data collected and a renewed interest in data-driven techniques such as machine learning (ML), has changed the landscape of analytics, and with it, time series forecasting. This book, filled with industry-tested tips and tricks, takes you beyond commonly used classical statistical methods such as ARIMA and introduces to you the latest techniques from the world of ML. This is a comprehensive guide to analyzing, visualizing, and creating state-of-the-art forecasting systems, complete with common topics such as ML and deep learning (DL) as well as rarely touched-upon topics such as global forecasting models, cross-validation strategies, and forecast metrics. You’ll begin by exploring the basics of data handling, data visualization, and classical statistical methods before moving on to ML and DL models for time series forecasting. This book takes you on a hands-on journey in which you’ll develop state-of-the-art ML (linear regression to gradient-boosted trees) and DL (feed-forward neural networks, LSTMs, and transformers) models on a real-world dataset along with exploring practical topics such as interpretability. By the end of this book, you’ll be able to build world-class time series forecasting systems and tackle problems in the real world.

Related Content you might be interested in

Current Title:

Small book image
Modern Time Series Forecasting with Python
Manu Joseph
Nov, 2022 | 552 pages
Small book image
Deep Learning for Time Series Cookbook
Luiacutes Roque | Vitor Cerqueira
Mar, 2024 | 274 pages
Small book image
Machine Learning for Time-Series with Python
Ben Auffarth
Oct, 2021 | 370 pages
Small book image
Practical Time Series Analysis
PKS Prakash | Avishek Pal
Sep, 2017 | 244 pages
Table of Contents (26 chapters)
Preface
Preface
Who this book is for
What this book covers
To get the most out of this book
Download the example code files
Download the color images
Conventions used
Get in touch
Share Your Thoughts
Download a free PDF copy of this book
1
Part 1 – Getting Familiar with Time Series
Part 1 – Getting Familiar with Time Series
Free Chapter
2
Chapter 1: Introducing Time Series
Chapter 1: Introducing Time Series
Technical requirements
What is a time series?
Data-generating process (DGP)
What can we forecast?
Forecasting terminology
Summary
Further reading
3
Chapter 2: Acquiring and Processing Time Series Data
Chapter 2: Acquiring and Processing Time Series Data
Technical requirements
Understanding the time series dataset
pandas datetime operations, indexing, and slicing 
– a refresher
Handling missing data
Mapping additional information
Saving and loading files to disk
Handling longer periods of missing data
Summary
4
Chapter 3: Analyzing and Visualizing Time Series Data
Chapter 3: Analyzing and Visualizing Time Series Data
Technical requirements
Components of a time series
Visualizing time series data
Decomposing a time series
Detecting and treating outliers
Summary
References
Further reading
5
Chapter 4: Setting a Strong Baseline Forecast
Chapter 4: Setting a Strong Baseline Forecast
Technical requirements
Setting up a test harness
Generating strong baseline forecasts
Assessing the forecastability of a time series
Summary
References
Further reading
6
Part 2 – Machine Learning for Time Series
Part 2 – Machine Learning for Time Series
7
Chapter 5: Time Series Forecasting as Regression
Chapter 5: Time Series Forecasting as Regression
Understanding the basics of machine learning
Time series forecasting as regression
Global forecasting models – a paradigm shift
Summary
References
Further reading
8
Chapter 6: Feature Engineering for Time Series Forecasting
Chapter 6: Feature Engineering for Time Series Forecasting
Technical requirements
Feature engineering
Avoiding data leakage
Setting a forecast horizon
Time delay embedding
Temporal embedding
Summary
9
Chapter 7: Target Transformations for Time Series Forecasting
Chapter 7: Target Transformations for Time Series Forecasting
Technical requirements
Handling non-stationarity in time series
Detecting and correcting for unit roots
Detecting and correcting for trends
Detecting and correcting for seasonality
Detecting and correcting for heteroscedasticity
AutoML approach to target transformation
Summary
References
Further reading
10
Chapter 8: Forecasting Time Series with Machine Learning Models
Chapter 8: Forecasting Time Series with Machine Learning Models
Technical requirements
Training and predicting with machine learning models
Generating single-step forecast baselines
Standardized code to train and evaluate machine learning models
Training and predicting for multiple households
Summary
References
Further reading
11
Chapter 9: Ensembling and Stacking
Chapter 9: Ensembling and Stacking
Technical requirements
Combining forecasts
Stacking or blending
Summary
References
Further reading
12
Chapter 10: Global Forecasting Models
Chapter 10: Global Forecasting Models
Technical requirements
Why Global Forecasting Models (GFMs)?
Creating GFMs
Strategies to improve GFMs
Bonus – interpretability
Summary
References
Further reading
13
Part 3 – Deep Learning for Time Series
Part 3 – Deep Learning for Time Series
14
Chapter 11: Introduction to Deep Learning
Chapter 11: Introduction to Deep Learning
Technical requirements
What is deep learning and why now?
Components of a deep learning system
Summary
References
Further reading
15
Chapter 12: Building Blocks of Deep Learning for Time Series
Chapter 12: Building Blocks of Deep Learning for Time Series
Technical requirements
Understanding the encoder-decoder paradigm
Feed-forward networks
Recurrent neural networks
Long short-term memory (LSTM) networks
Gated recurrent unit (GRU)
Convolution networks
Summary
References
Further reading
16
Chapter 13: Common Modeling Patterns for Time Series
Chapter 13: Common Modeling Patterns for Time Series
Technical requirements
Tabular regression
Single-step-ahead recurrent neural networks
Sequence-to-sequence (Seq2Seq) models
Summary
Reference
Further reading
17
Chapter 14: Attention and Transformers for Time Series
Chapter 14: Attention and Transformers for Time Series
Technical requirements
What is attention?
The generalized attention model
Forecasting with sequence-to-sequence models and attention
Transformers – Attention is all you need
Forecasting with Transformers
Summary
References
Further reading
18
Chapter 15: Strategies for Global Deep Learning Forecasting Models
Chapter 15: Strategies for Global Deep Learning Forecasting Models
Technical requirements
Creating global deep learning forecasting models
Using time-varying information
Using static/meta information
Using the scale of the time series
Balancing the sampling procedure
Summary
Further reading
19
Chapter 16: Specialized Deep Learning Architectures for Forecasting
Chapter 16: Specialized Deep Learning Architectures for Forecasting
Technical requirements
The need for specialized architectures
Neural Basis Expansion Analysis for Interpretable Time Series Forecasting (N-BEATS)
Neural Basis Expansion Analysis for Interpretable Time Series Forecasting with Exogenous Variables (N-BEATSx)
Neural Hierarchical Interpolation for Time Series Forecasting (N-HiTS)
Informer
Autoformer
Temporal Fusion Transformer (TFT)
Interpretability
Probabilistic forecasting
Summary
References
Further reading
20
Part 4 – Mechanics of Forecasting
Part 4 – Mechanics of Forecasting
21
Chapter 17: Multi-Step Forecasting
Chapter 17: Multi-Step Forecasting
Why multi-step forecasting?
Recursive strategy
Direct strategy
Joint strategy
Hybrid strategies
How to choose a multi-step forecasting strategy?
Summary
References
22
Chapter 18: Evaluating Forecasts – Forecast Metrics
Chapter 18: Evaluating Forecasts – Forecast Metrics
Technical requirements
Taxonomy of forecast error measures
Investigating the error measures
Experimental study of the error measures
Guidelines for choosing a metric
Summary
References
Further reading
23
Chapter 19: Evaluating Forecasts – Validation Strategies
Chapter 19: Evaluating Forecasts – Validation Strategies
Technical requirements
Model validation
Holdout strategies
Cross-validation strategies
Choosing a validation strategy
Validation strategies for datasets with multiple time series
Summary
References
Further reading
24
Index
Index
Why subscribe?
25
Other Books You May Enjoy
Other Books You May Enjoy
Packt is searching for authors like you
Share Your Thoughts
Download a free PDF copy of this book

To get the most out of this book

You should have basic familiarity with Python programming, as the entire code that we use for the practical sections is in Python. Familiarity with major libraries in Python, such as pandas and scikit-learn, are not essential (because the book covers some basics) but will help you get through the book much faster. Familiarity with PyTorch, the framework the book uses for deep learning, is also not essential but would accelerate your learning by many folds. Any of the software requirements shouldn’t stop you because, in today’s internet-enabled world, the only thing that is standing between you and a world of knowledge is the search bar in your favorite search engine.

Another key aspect to get the most out of this book is to run the associated notebooks as you go along the lessons. Also, feel free to experiment with different variations that the book doesn’t go into. That is a surefire way to internalize what’s being talked about in the book. And for that, we need to set up an environment, as you’ll see in the following section.

Setting up an environment

The easiest way to set up an environment is by using Anaconda, a distribution of Python for scientific computing. You can use Miniconda, a minimal installer for Conda, as well if you do not want the pre-installed packages that come with Anaconda:

  1. Install Anaconda/Miniconda: Anaconda can be installed from https://www.anaconda.com/products/distribution. Depending on your operating system, choose the corresponding file and follow the instructions. Alternatively, you can install Miniconda from here: https://docs.conda.io/en/latest/miniconda.html#latest-miniconda-installer-links.
  2. Open conda prompt: To open Anaconda Prompt (or Terminal on Linux or macOS), do the following:
    • Windows: Open the Anaconda Prompt (Start | Anaconda Prompt)
    • macOS: Open Launchpad and then open Terminal. Type conda activate.
    • Linux: Open Terminal. Type conda activate.
  3. Navigate to the downloaded code: Use operating system-specific commands to navigate to the folder where you have downloaded the code. For instance, in Windows, use cd.
  4. Install the environment: Using the anaconda_env.yml file that is included, install the environment:
    conda env create -f anaconda_env.yml

This creates a new environment under the name modern_ts and will install all the required libraries in the environment. This can take a while.

  1. Checking the installation: We can check whether all the libraries required for the book are installed properly by executing a script in the downloaded code folder:
    python test_installation.py
  2. Activating the environment and running notebooks: Every time you want to run the notebooks, first activate the environment using the conda activate modern_ts command and then use the Jupyter Notebook (jupyter notebook) or JupyterLab (jupyter lab), according to your preference.

Download the data

You are going to be using a single dataset throughout the book. The book uses the London Smart Meters dataset from Kaggle for this purpose. Therefore, if you don’t have an account with Kaggle, please go ahead and create one: https://www.kaggle.com/account/login?phase=startRegisterTab.

There are two ways you can download the data-automated and manual.

For the automated way, we need to download a key from Kaggle. Let’s do that first (if you are going to choose the manual way, you can skip this):

  1. Click on your profile picture in the top-right corner of Kaggle.
  2. Select Account, and find the section for API.
  3. Click the Create New API Token button. A file with the name kaggle.json will be downloaded.
  4. Copy the file and place it in the api_keys folder in the downloaded code folder.

Now that we have kaggle.json downloaded and placed in the right folder, let’s look at the two methods to download data:

Method one – automated download

  1. Activate the environment using conda activate modern_ts.
  2. Run the provided script from the root directory of the downloaded code:
    python scripts/download_data.py

That’s it. Now, just wait for the script to finish downloading, unzip it, and organize the files in the expected format.

Method two – manual download

  1. Go to https://www.kaggle.com/jeanmidev/smart-meters-in-london and download the dataset.
  2. Unzip the contents to data/london_smart_meters.
  3. Unzip hhblock_dataset to get the raw files we want to work with.
  4. Make sure the unzipped files are in the expected folder structure (see the next section).

Now that you have downloaded the data, we need to make sure it is arranged in the following folder structure. The automated download does it automatically, but with the manual download, this structure needs to be created. To avoid ambiguity, the expected folder structure can be found as follows:

data
├── london_smart_meters
│   ├── hhblock_dataset
│   │   ├── hhblock_dataset
│   │       ├── block_0.csv
│   │       ├── block_1.csv
│   │       ├── ...
│   │       ├── block_109.csv
│── acorn_details.csv
├── informations_households.csv
├── uk_bank_holidays.csv
├── weather_daily_darksky.csv
├── weather_hourly_darksky.csv

There can be additional files as part of the extraction process. You can remove them without impacting anything. There is a helpful script that checks this structure.

python test_data_download.py

If you are using the digital version of this book, we advise you to type the code yourself or access the code from the book’s GitHub repository. Doing so will help you avoid any potential errors related to the copying and pasting of code.

The code that is provided along with the book is in no way a library but more of a guide for you to start experimenting on. The amount of learning you can derive from the book and code is directly proportional to how much you experiment with the code and stray outside your comfort zone. So, go ahead and start experimenting and putting the skills you pick up in the book to good use.

Previous Section
Next Section