This document provides an introduction to machine learning and its applications in aviation. It begins with definitions of key machine learning concepts like supervised learning, unsupervised learning, and reinforcement learning. It then discusses popular machine learning algorithms and how they are applied in areas like predictive maintenance, flight optimization, and customer analytics. The document also introduces RapidMiner, an open-source platform for machine learning projects, and provides an overview of its graphical user interface and basic terminology.

1. This course is prepared under the Erasmus+ KA-210-YOU Project titled
«Skilling Youth for the Next Generation Air Transport Management»
Machine Learning
Applications in Aviation
Introduction to Machine Learning and RapidMiner
Asst. Prof. Dr. Emircan Özdemir
Eskişehir Technical University

2. • Machine learning is the dynamic field of artificial intelligence that empowers
computers to learn and improve from experience without being explicitly
programmed.
• It's about enabling machines to recognize patterns, make predictions, and
optimize decisions based on data.
• In the aviation industry, machine learning serves as a catalyst for transformative
advancements, revolutionizing areas such as predictive maintenance and
enabling highly efficient flight path optimization. Furthermore, it empowers air
transport companies to gain deeper insights into passenger behavior,
preferences, and needs, enhancing marketing strategies to provide passengers
with unparalleled experiences.
Machine Learning Applications in Aviation 2
What is Machine Learning?

3. • By extracting valuable insights from vast aviation datasets, machine
learning is instrumental in making aviation safer, more efficient, and cost-
effective.
• As we progress through this course, you'll gain a deeper understanding of
how machine learning is harnessed to tackle real-world challenges in
aviation.
• But first, it’s important to establish a clear understanding of how machine
learning, artificial intelligence, deep learning, data science, big data,
and data mining are interconnected.
Machine Learning Applications in Aviation 3
What is Machine Learning (ML)?

4. • Artificial Intelligence (AI): AI is the field of computer science and technology that
focuses on creating computer systems and machines capable of performing tasks that
typically require human intelligence. These tasks include learning from experience,
reasoning, problem-solving, understanding natural language, and perceiving and
interacting with the environment.
AI aims to develop machines that can simulate and replicate human cognitive functions,
enabling them to adapt and excel in various domains, from image recognition and language
processing to autonomous decision-making.
• Machine Learning (ML): ML is a subset of artificial intelligence that involves the
development of algorithms and models that enable computer systems to learn from and
make predictions or decisions based on data.
Machine Learning Applications in Aviation 4
Definition of Terms and Fields Related with ML

5. In essence, ML is the practice of training a machine
to recognize patterns, gain insights, and improve
its performance on a specific task without being
explicitly programmed for that task.
• Deep Learning (DL): Deep learning is a subfield of machine learning that focuses on the use
of artificial neural networks, particularly deep neural networks, to model and solve complex
tasks. These deep neural networks consist of multiple layers of interconnected nodes
(neurons) that are capable of learning and representing intricate patterns and features from
data.
DL has demonstrated remarkable success in tasks such as image and speech recognition,
natural language processing, and many other areas that involve complex, unstructured data. It
excels in handling large, high-dimensional datasets and is a key technology driving advances in
artificial intelligence.
Machine Learning Applications in Aviation 5
Definition of Terms and Fields Related with ML
Source: https://medium.com/codex/classical-programming-vs-machine-learning-in-plain-english-3f39c56673d9

6. Machine Learning Applications in Aviation 6
Relationship Between AI, ML and DL
Source: https://medium.com/t%C3%BCrkiye/makine-%C3%B6%C4%9Frenmesi-nedir-20dee450b56e

7. • Big Data: Big data refers to extremely large and complex datasets that exceed the
capabilities of traditional data processing and analysis methods. Big data is characterized
by its volume, velocity, variety, and veracity. Big data often includes massive amounts of
structured and unstructured data from various sources, such as social media, sensors,
and transaction records.
Examples of structured and unstructured data in aviation management:
Structured data: Flight Schedules, Passenger Data, Maintenance Records, Financial Data,
Crew Schedules (organized and formatted with a clear and predefined structure)
Unstructured data: Customer Feedback and Reviews, Social Media Posts, Safety
Reports, Email Correspondence, Voice Recordings, Weather Data (lacks a specific format
or structure)
Machine Learning Applications in Aviation 7
Definition of Terms and Fields Related with ML

8. • Data Science: Data science is a multidisciplinary field that combines various techniques,
including statistics, data analysis, machine learning, and domain knowledge, to
extract knowledge and insights from data. It involves the entire data analysis process,
from data collection and cleaning to modeling and interpretation.
Data scientists use their expertise to formulate hypotheses, design experiments, and apply
advanced analytics to discover patterns and make informed decisions. Data science plays a
crucial role in a wide range of applications, from business intelligence to scientific research.
Machine Learning Applications in Aviation 8
Definition of Terms and Fields Related with ML

9. Machine Learning Applications in Aviation 9
Relationship Between AI, ML,DL and Data Science
Source: Kotu, Vijay, and Bala Deshpande. 2019. “Chapter 1 - Introduction.” In , edited by Vijay Kotu and Bala B T - Data Science (Second Edition) Deshpande, 1–18. Morgan Kaufmann.
https://doi.org/https://doi.org/10.1016/B978-0-12-814761-0.00001-0.

10. • Data Mining: Data mining is a specific subset of data science that focuses on discovering
patterns, trends, and valuable information within large datasets. It involves the use of
various techniques, such as clustering, association rule mining, and regression, to
unearth hidden insights. These techniques may include ML, traditional statistical methods,
exploratory data analysis, and other data analysis techniques.
Data mining is often applied to support decision-making and predict future trends. It
has applications in areas like marketing, customer relationship management, fraud
detection, and more. Data mining can be a vital component of data science, as it involves
uncovering valuable nuggets of knowledge within data.
Data mining framework and process will be discussed further in the next chapter.
Machine Learning Applications in Aviation 10
Definition of Terms and Fields Related with ML

11. • There is also one last important term,
which is data analytics.
• Data analytics is the process of
examining, cleaning, transforming, and
modeling data to discover valuable
insights, draw conclusions, and support
decision-making. It involves applying
various techniques and tools to explore
and interpret data, identify trends,
patterns, and insights, and make data-
driven recommendations or predictions.
• In summary, data analytics is a broader
field that includes data mining as one of
its methodologies.
Machine Learning Applications in Aviation 11
Definition of Terms and Fields Related with ML

12. • This course introduces the most important machine learning algorithms and data mining
techniques to enable you to use them in real-world air transport applications.
• Machine learning algorithms are fundamentally divided into three groups. These are
supervised learning, unsupervised learning and reinforcement learning.
• Supervised Learning: Algorithms are trained on labeled data, where each input is
paired with the correct output. The goal is to learn a mapping from inputs to outputs,
enabling the model to make predictions or classifications on new, unseen data.
• Common supervised learning tasks include regression (predicting numerical values) and
classification (categorizing data into classes or labels).
• Examples of supervised learning algortihms: Linear Regression, Logistic Regression,
Support Vector Machines, Decision Trees, Random Forest, Neural Networks, Naive
Bayes, K-Nearest Neighbors (KNN).
Machine Learning Applications in Aviation 12
Machine Learning

13. • Unsupervised learning: Unsupervised learning deals with unlabeled data, and the
objective is to discover patterns, structures, or groupings within the data. In unsupervised
learning, the algorithm learns on its own without any guidance based on known outcomes.
• Common unsupervised learning tasks include clustering (grouping similar data points),
dimensionality reduction (reducing the number of features), and density estimation.
• Examples of unsupervised learning algorithms: K-Means Clustering, Principal Component
Analysis (PCA), Singular Value Decomposition, Hierarchical Clustering, Apriori Algorithm.
Machine Learning Applications in Aviation 13
Machine Learning
Source:
https://eneshazr.medium.com/supervised-
unsupervised-learning-makine-
%C3%B6%C4%9Frenmesi-
b903bc09430e

14. • Reinforcement learning: It is used for sequential decision-making problems where an
agent interacts with an environment. The agent learns to make a sequence of
decisions to maximize a cumulative reward signal.
• This type of learning is often used in robotics, autonomous systems, and game playing.
• Examples of reinforcement learning algorithms: Q-Learning, Deep Q-Networks (DQN),
and Proximal Policy Optimization (PPO).
Machine Learning Applications in Aviation 14
Machine Learning
Source:
https://techvidvan.com/tutorials/reinforcement-
learning/

15. Machine Learning Applications in Aviation 15
Source: https://towardsdatascience.com/

16. Machine Learning Applications in Aviation 16
Advantages and Disadvantages of Supervised and
Unsupervised Learning
Source: Caballé-Cervigón, Nuria, José L. Castillo-Sequera, Juan A. Gómez-Pulido, José M. Gómez-Pulido, and María L. Polo-Luque. 2020.
“Machine Learning Applied to Diagnosis of Human Diseases: A Systematic Review.” Applied Sciences 10 (15): 5135. https://doi.org/10.3390/app10155135.

17. • These three fundamental groups of machine learning algorithms each have their own set
of techniques and applications. Additionally, many other algorithms and methods exist,
often combining elements from these basic groups. For example, semi-supervised
learning uses a mix of labeled and unlabeled data, and deep learning employs neural
networks to model complex patterns and relationships.
• In this course, we will explore the rich landscape of machine learning algorithms, with a
strong focus on both supervised and unsupervised learning techniques. These two
categories are common and foundational in the field of machine learning.
Machine Learning Applications in Aviation 17
Machine Learning

18. Data Collection
Data
Preprocessing
Feature
Engineering
Model
Selection
Model Training
Model
Evaluation
Model
Optimization
Interpretation
and
Visualization
Deployment
Machine Learning Applications in Aviation 18
Machine Learning Process

19. 1- Data Collection: Gathering relevant data from various sources, which may include
structured data from databases, text documents, or other data types, depending on the
problem.
2- Data Preprocessing: Cleaning, transforming, and organizing the data to make it suitable
for analysis. This step often involves handling missing values, dealing with outliers, and
scaling or normalizing features.
3- Feature Engineering: Selecting, creating, or transforming features to extract the most
relevant information from the data. This step can significantly impact the performance of
machine learning models.
4- Model Selection: Choosing an appropriate machine learning algorithm or model based
on the problem type (supervised or unsupervised), the nature of the data, and the specific
goals of the project.
Machine Learning Applications in Aviation 19
Machine Learning Process

20. 5- Model Training: Training the selected machine learning model on the preprocessed
data. In supervised learning, this involves using labeled data to learn patterns and
relationships, while in unsupervised learning, it's about discovering inherent structures in
the data.
6- Model Evaluation: Assessing the performance of the trained model using appropriate
evaluation metrics. The choice of metrics depends on the type of problem, such as
accuracy, precision, recall, F1 score, or clustering metrics like Silhouette Score.
7- Model Optimization: Fine-tuning the model and its hyperparameters to improve its
performance. This may involve techniques like cross-validation and grid search.
8- Interpretation and Visualization: Examining the model's results, visualizing findings,
and gaining insights from the analysis. This step is crucial for understanding the meaning
behind the patterns or predictions.
9- Deployment: If the machine learning model is intended for real-world use, it may be
deployed in a production environment to make predictions or decisions.
Machine Learning Applications in Aviation 20
Machine Learning Process

21. RapidMiner is a data science and machine learning platform that provides tools and
features for data preparation, machine learning, and predictive analytics. It is designed to
help organizations and data professionals with various tasks related to data analysis,
modeling, and decision-making.
RapidMiner provides data mining and machine learning procedures including: data loading
and transformation (ETL), data preprocessing and visualization, predictive analytics and
statistical modeling, evaluation, and deployment. RapidMiner is written in the Java
programming language.
RapidMiner has over 1 million users globally and is used in a wide range of industries and
applications, from marketing and finance to healthcare and manufacturing, to extract
valuable insights and make data-driven decisions. It is known for its user-friendly interface
and versatility in data science and analytics workflows.
Machine Learning Applications in Aviation 21
Getting Started with Rapidminer
Source: https://rapidminer.com/

22. There are 4,000+ universities around the world using RapidMiner every day in their data
science programs. RapidMiner Educational License Program provides free, 1-year,
renewable educational licenses for RapidMiner Studio.
If you are a student at an accredited university, you qualify for a 1-year, renewable
educational license of RapidMiner Studio.
For registering an educational license and downloading RapidMiner Studio please visit:
https://rapidminer.com/platform/educational/
Machine Learning Applications in Aviation 22
Rapidminer Educational License Program

23. RapidMiner Studio is the GUI-based software where data mining and predictive analytics
workflows can be built and deployed.
You can do almost everything in data science and data mining without writing a single line
of code using RapidMiner.
Read the article in the link below:
https://www.kdnuggets.com/drag-drop-analyze-the-rise-of-nocode-data-science
Machine Learning Applications in Aviation 23
Getting Started with Rapidminer

24. • RapidMiner provides a user-friendly graphical user interface (GUI) that allows users to
perform a wide range of data preparation, modeling, and analysis tasks.
Machine Learning Applications in Aviation 24
GUI of the Rapidminer

25. • You can activate diferent views selecting the View -> Show Panel menu option.
Machine Learning Applications in Aviation 25
GUI of the Rapidminer

26. 1- Design Panel: This is the central workspace where you create and edit your data
science workflows. You can drag and drop operators (data preprocessing, modeling,
analysis, etc.) onto the design panel and connect them to define the workflow.
2- Operators: These are the building blocks of your workflows. Operators represent specific
actions, such as data loading, data preprocessing, modeling, and evaluation. You can find
and add operators from the Operator Toolbox.
3- Operator Toolbox: Located on the left side of the GUI, the Operator Toolbox provides a
wide range of operators categorized by functionality. You can browse and search for
operators to add to your workflows.
4- Results Panel: This panel displays the results of your analysis, including visualizations,
performance metrics, and data previews. You can see the output of each operator as you
execute your workflow.
Machine Learning Applications in Aviation 26
Main Components of the Rapidminer GUI

27. 5- Repository Panel: The repository is where you can organize and store various
components, such as workflows, data, models, and templates. You can manage, search for,
and reuse items from the repository.
6- Process Panel: The process panel provides an overview of your workflow and allows
you to configure the properties of selected operators. You can set parameters, adjust
settings, and view operator documentation.
7- Logging and Errors: At the bottom of the GUI, you can view log messages and error
notifications. This helps you track the progress of your workflow and troubleshoot any
issues.
8- Menus and Toolbar: RapidMiner includes various menus and a toolbar at the top of the
GUI for actions like opening, saving, running workflows, and accessing additional features
and options.
Machine Learning Applications in Aviation 27
Main Components of the Rapidminer GUI

28. 9- Control Flow: You can define control flow within workflows by using control operators
like loops, branches, and conditions. This allows for more complex workflow design.
10- Connectors: These lines between operators represent the data flow in your workflow.
You connect operators to specify how data is passed from one step to the next.
Machine Learning Applications in Aviation 28
Main Components of the Rapidminer GUI

29. A typical process in Rapidminer
consists of several operators.
Example is given for a decision
tree model, reading data from
Excel files.
Machine Learning Applications in Aviation 29

30. Example (Data Point): An example, also known as a data point or observation, represents
a single instance or record in your dataset. It can be a row in a table, a data point in a
scatterplot, or any individual data entity.
Attributes (Features): Attributes, often referred to as features or variables, are the
characteristics or properties that describe each example. They can be numeric, categorical,
or text-based, and they serve as the input to your data analysis.
Operator: Operators are building blocks in RapidMiner workflows that perform specific data
operations or actions. They can include data preprocessing operators, modeling operators,
and evaluation operators, among others.
Process: A process, or workflow, is a visual representation of a series of connected
operators that define a data analysis task. It describes the sequence of operations
performed on your data.
Machine Learning Applications in Aviation 30
Terminology of the Rapidminer

31. Machine Learning Applications in Aviation 31

32. • There are many options to read (import) and write (export) data in RapidMiner. You can
find them in the Operators panel and use the search box.
• Some of data importing tools:
Read Database, Read Excel, Read CSV, Read XML,
Read Document, Read URL, Read BibTeX, Read XML,
Read Access etc.
• Some of data exporting tools:
Write Database, Write Excel, Write CSV, Write Access,
Write Document etc.
Machine Learning Applications in Aviation 32
Data Importing and Exporting Tools

33. Also you can;
• …store your data using the Store operator.
This will allow you to store your database at a location
in the data repository as a RapidMiner IO Object (IO: input-output).
IO Object refers to an object or data structure used to represent
input (read) and output (write) data connections in a data science or
machine learning workflow. It can refer also models in addition to
databases.
• …retrieve an existing data using the Retrieve operator.
• …use Cloud Storage operators to keep/update your data
in cloud services/platforms.
Machine Learning Applications in Aviation 33
Data Importing and Exporting Tools

34. Select the Read Excel operator
Machine Learning Applications in Aviation 34
Example: Importing Excel Data

35. Click on the «Import Configuration Wizard» button and then select the file path of
data_golf.xlsx file that you can download from the lesson materials section below.
Machine Learning Applications in Aviation 35
Example: Importing Excel Data

36. Select the cells that you want to import.
Or you can simply press Select All.
Also please define the header row
as attribute names.
Then, press Next button.
Machine Learning Applications in Aviation 36
Example: Importing Excel Data

37. In the next window, you can change data types,
variable roles, rename variables, or exclude
variables.
After formatting your varibles,
press the Next button.
Machine Learning Applications in Aviation 37
Example: Importing Excel Data

38. Using the parameters of the Store operator, define a repository entry name for storing data
in your RapidMiner repository.
Machine Learning Applications in Aviation 38
Example: Importing Excel Data

39. After creating your data import process, press the execute process button on the top left.
Or you can use F11 shortcut key to run the process.
After running the process, your data will be stored into the RapidMiner repository.
Next time you can call your data from the list in the repository.
Machine Learning Applications in Aviation 39
Example: Importing Excel Data

40. • Within this course, we will follow our subjects using RapidMiner. Therefore, the following
links are provided to you for exploring more features of the RapidMiner.
For data import: https://academy.rapidminer.com/learn/video/importing-data-in-rapidminer-
studio
For data types and conversions: https://rapidminer.com/blog/data-prep-data-types-
conversions/
RapidMiner Academy: https://academy.rapidminer.com/
RapidMiner Blog: https://rapidminer.com/blog/
Machine Learning Applications in Aviation 40
Conclusion