Course:

«Selection of simple algorithms for Machine Learning applications»

Lesson 2: Classifying Data Patterns.

Learning Spaces and Digital Pedagogy – Winter 2022

Author: Juan Jose Martinez Castillo

*Note: based in «Selection of simple algorithms for Machine Learning applications», Learning Spaces and Digital Pedagogy – Winter 2022, University of Calgary, Canada.

Overview

Include an overview of the lesson

What context will the lesson be taught in, what type of learners is the lesson for?

These lessons are part of an initial and basic training course in machine learning. The students have already had to pass an introductory course in mathematics and basic concepts related to machine learning.

What happens in the lesson?

In these educational lessons, the student acquires the ability to recognize patterns and which machine learning algorithms are the most suitable to get the most out of the data, represented in those patterns.

How do students use the digital artifact as part of the lesson?

These lessons, in the form of online tutorials, use WordPress, and its communication forums. It is also proposed to the student the use of drawing software to create their patterns, but if they take the alternative of «doing them by hand», they will still have to digitize them. In both situations, you must insert the image of your work, or embed the internet link, in your comments. In lesson 2, you are given a task to perform using a Powerpoint template, open it, modify it, identify it with your data, capture a screen image, and insert the image into your wordpress forum comments. Also, in a more advanced option, an online Google template could be used, which the student would modify.

Outcomes

Provide 1-2 learning outcomes that describe what the outcomes for the lesson are for the learners

The student will develop the ability to recognize when there is image or abstract (numerical) patterns.
The student will learn to select the machine learning algorithm that best suits the available data pattern.

Assessment

Describe how an instructor sees if learners are achieving the learning objectives. Is the assessment summative, as a grade, or formative, as a way of gauging learning? How can learners act on the assessment?

The student must perform tasks that must be evaluated by the instructor, at the same time, that he must explain and support his work with arguments. The student will not be able to advance to the next lesson until he gets the answer from his instructor. The student must perform tasks that must be evaluated by the instructor, at the same time, that he must explain and support his work with arguments. The student will not be able to advance to the next lesson until he gets the answer from his instructor.

Rationale

Provide a description of why you chose to design the lesson as you did and how the digital artifact supports the lesson and the learning outcomes. Base your rationale on what you’ve learned in LSDP and your research into pedagogy in the lesson’s field. Make sure your rational cites relevant literature.

The use of computational thinking, and the understanding of everything related to the search for patterns in data, is essential for the development of skills in students of basic and advanced machine learning courses. In the lessons, an order is established, step by step, in an algorithmic way, so that the student becomes familiar with working in this way, since in advanced courses, they must develop work «pipelines» to execute complex processes using Big Data, and cloud computing, where “routine” tasks must be carefully performed. A student of Artificial Intelligence, and specifically of machine learning, must have a broad development of her abstract reasoning skills; and this aspect goes beyond the simple understanding of an algorithm, but rather, in the permanent search for «patterns» that allow it not only to select the best algorithm for that particular case, but also to be able to modify it to optimize its results and minimize the error and the possible bias that may have with the data collected; We must not forget that, in the end, machine learning is nothing more than making a logical decision, based on data, or trying to «predict» future behavior, precisely, based on behavior patterns. For this reason, I have tried to cover both aspects: thinking step by step (like an algorithm) and teaching what a «pattern» is, from simple pencil lines to apparently complex groups of data labeled in some way, to be able to be used.

I used The SAMR Model for integrating technology into teaching:

•Substitution: Students read an online article discussing about machine learning (replacing notebooks). Students Use of WordPress, and presentation software (like PowerPoint or Prezi) to construct a presentation providing information about a selected locale.

•Augmentation: Incorporate interactive multimedia, audio, video, hyperlinks in the presentation (website) to give more depth and provide more engaging presentation.

•Modification: Using a PowerPoint template, Students drag and drop on top of the grouped data they think is the best machine learning algorithm, for that case, and they explain why you made that selection.

•Redefinition: Students watch images / Gif examples, learn and practice the techniques step by step (Computing thinking and solving problems), then the instructor provides feedback about their work.

References

What is machine learning?
https://www.techtarget.com/searchenterpriseai/definition/machine-learning-ML

What is Computational Thinking?
https://digitalpromise.org/initiative/computational-thinking/computational-thinking-for-next-generation-science/what-is-computational-thinking/

What is Computational Thinking? Why thinking like a computer builds skills for success.
https://teachyourkidscode.com/what-is-computational-thinking/

Abstract Thinking: What It Is, Why We Need It, and When to Rein It In.
https://www.healthline.com/health/abstract-thinking

Big Data and the Importance of pattern discovery.
https://datapeaker.com/big-data/reconocimiento-de-patrones-importancia-del-reconocimiento-de-patrones/

The SAMR model: A Powerful Model for Understanding Good Tech Integration.
https://www.edutopia.org/article/powerful-model-understanding-good-tech-integration

Examples of Applying the SAMR Model can Help Teachers Understand and Embrace it.
https://www.emergingedtech.com/2015/04/examples-of-transforming-lessons-through-samr/