IntroML_1_Introduction_Tagged.pdf
0 likes48 views
This document outlines the course details for an Introduction to Machine Learning module. The module will cover the basics of machine learning including supervised learning techniques like classification and regression, as well as unsupervised learning techniques like clustering and dimensionality reduction. Students will learn to prepare data, visualize it, evaluate models, and communicate results to both technical and non-technical audiences. The course will involve lectures, labs, homework assignments, and a final written test. The goal is for students to understand the basics of machine learning and be able to apply the techniques to analyze real-world data.
![Trinity College Dublin, The University of Dublin 18
What is machine learning (ML)?
Quote Slide Option 2
Instead of trying to produce a programme to simulate the adult mind, why not
rather try to produce one which simulates the child's? [Alan Turing, 1950]](https://image.slidesharecdn.com/introml1introduction-tagged-230519105826-0707f325/85/IntroML_1_Introduction_Tagged-pdf-18-320.jpg)
![Trinity College Dublin, The University of Dublin 25
What can we do with ML?
Quote Slide Option 2
“Hands-On Machine Learning with Scikit-Learn,
Keras, and TensorFlow”, Aurélien Géron, 2019
In the simplest case, the ML model learns just once. It will not learn later on or change automatically.
So, it learns once and then it is used as an anomaly detection tool.
X = [feature1, feature2, …, featurek]](https://image.slidesharecdn.com/introml1introduction-tagged-230519105826-0707f325/85/IntroML_1_Introduction_Tagged-pdf-25-320.jpg)
IntroML_1_Introduction_Tagged.pdf
- 1. INSERT IMAGE HERE Introduction to Machine Learning (5 ECTS) Giovanni Di Liberto Asst. Prof. in Intelligent Systems, SCSS Room G.15, O’Reilly Institute ©Trinity College Dublin
- 2. Trinity College Dublin, The University of Dublin Quote Slide Option 2 Lecture outline • Module outline and schedule • Module objectives and evaluation criteria • What is Machine Learning (ML) • What can we do with ML
- 3. Trinity College Dublin, The University of Dublin Quote Slide Option 2 Course outline • (Introduction to programming – Python) • Data visualisation • Supervised learning • Classification • Regression and time-series • Overfitting and regularisation • Unsupervised learning • Clustering • Dimensionality reduction • Model tuning, evaluation, and Feature selection • Tutorials + continuous assessment on lab exercises • Examples on publicly available data • Discussion about current and future challenges for ML • E.g., data sharing, anonymisation and privacy, standardisation
- 4. Trinity College Dublin, The University of Dublin 4 Week LECTURE LAB Additional hour and deadlines 1 Overview Introduction to programming and Python programming tutorial 2 Descriptive stats vs. ML. Data visual. Discussion board task explanation Python programming and basic visualisation tutorial 3 Supervised Learn. Simple classifiers Data visualisation and simple classification tutorial 4 Classification – part 1 Sup.L. Model quality evaluation tutorial Data preparation and visualization test (1h) – immediately after the tutorial 5 Classification – part 2: algorithms Homework 1 explanation Classification tutorial 6 Classification part 3 and Regression Regression and time-series tutorial 7 Reading week - 8 Regression Homework 2 explanation Unsupervised learning tutorial Homework 1 deadline 9 Regression and Unsupervised learning Homework hour with Q&A 10 Recap and feature selection Anomaly detection tutorial 11 Data sharing, storage, and privacy Homework hour with Q&A Homework 2 deadline 12 Guest lecture Discussion Introduction to Machine Learning – 2022-2023 Written test (2h)
- 5. Trinity College Dublin, The University of Dublin 5 Evaluation Quote Slide Option 2 • Laboratory test (10%) • Individual homework (25%) • Group assignment (25%) • Written test (40%) Theory Technical skills Communication 5 ECTS = 125h
- 6. Trinity College Dublin, The University of Dublin 6 Useful references Quote Slide Option 2 • Python tutorial https://coherentpdf.com/python/pythonfromtheverybeginning.html • Jupyter-notebook tutorial https://www.dataquest.io/blog/jupyter-notebook-tutorial/ “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019
- 7. Trinity College Dublin, The University of Dublin 7 Introduction to ML Quote Slide Option 2 - ML (what is it? How does it work? How do we use it?) - Most common algorithms - Challenges in the context of smart-cities Theory Technical skills Communication - Introduction to computer programming - Python - Data visualization - Data analysis - ML - The concept/algorithm matters more than the syntax or issues with libraries! - Expert – non-expert interaction - Problem -> specify requirements -> understand challenges -> interpret results
- 8. Trinity College Dublin, The University of Dublin 8 Introduction to ML Quote Slide Option 2 Theory Technical skills Communication - This module will give you the basics. A good start, but you won’t be ML experts (but you can get there!) - You will learn how to communicate with ML experts - It is essential that you play around with the code, try different variations on other datasets. Re-running our tutorials won’t be enough - Interact with us! Ask questions. Use the online resources. Don’t wait, start straight away!
- 9. Trinity College Dublin, The University of Dublin 9 Introduction to ML Quote Slide Option 2 - Issues with the code during the assessment/labs? No problem. Ask us questions or use pseudocode to describe exactly what you wanted to do Use the Internet!
- 10. Trinity College Dublin, The University of Dublin 10 Introduction to ML Quote Slide Option 2 - Let’s not reinvent the wheel! - We will study what tools are available, how they work, and how to best use them! - But not black-box style!
- 11. Trinity College Dublin, The University of Dublin 11 Your Continuous Feedback is Important! Quote Slide Option 2
- 12. Trinity College Dublin, The University of Dublin 12 Your Background and Prior Experience Quote Slide Option 2 • Laptop/computer • Excel • OS (Operating System) • Coding (Python) • What’s a variable • Integer, double, logical • for loop, function, class? • What’s a compiler?
- 13. Trinity College Dublin, The University of Dublin 13 https://jupyter.readthedocs.io/en/latest/install/notebook-classic.html Setting up your coding environment Download and install Anaconda - Windows or Mac OS: run Anaconda Navigator from the Start menu or application menu - In Linux: run anaconda-navigator from the terminal
- 14. Trinity College Dublin, The University of Dublin 14 What is Artificial Intelligence (AI)? Quote Slide Option 2
- 15. Trinity College Dublin, The University of Dublin 15 What is Artificial Intelligence (AI)? Quote Slide Option 2 https://www.youtube.com/watch?v=760lA2YCKjM https://www.youtube.com/watch?v=78-1MlkxyqI
- 16. Trinity College Dublin, The University of Dublin 16 What is Artificial Intelligence (AI)? Quote Slide Option 2 Weak AI Strong AI “Artificial intelligence: A modern approach”, Russel and Norvig • The study of intelligent agents • Systems/devices that perceive their environment and take actions in that environment to achieve their goals • Weak AI: “Intelligent” actions (like or better than humans) • Strong AI: “Intelligent” thinking
- 17. Trinity College Dublin, The University of Dublin 17 What is machine learning (ML)? Quote Slide Option 2 AI ML “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019 Learning can be seen as a process for improving performance based on experience To be defined To be defined
- 18. Trinity College Dublin, The University of Dublin 18 What is machine learning (ML)? Quote Slide Option 2 Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child's? [Alan Turing, 1950]
- 19. Trinity College Dublin, The University of Dublin 19 What is machine learning (ML)? Quote Slide Option 2 “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019
- 20. Trinity College Dublin, The University of Dublin 20 What can we do with ML? Quote Slide Option 2 “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019 ML is generally about learning patterns in the data and use that information for our goals (e.g., spam filter) How can we learn these patterns? We don’t tell the algorithm how to detect spam emails. We give the ML algorithm examples of spam emails. Then, it has to figure out how to detect them.
- 21. Trinity College Dublin, The University of Dublin 21 What can we do with ML? Quote Slide Option 2 Supervised learning • Classification • Regression Unsupervised learning “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019
- 22. Trinity College Dublin, The University of Dublin 22 Supervised Learning Quote Slide Option 2 y = f(X) f ynew Model Training (learning or fit) Xnew f y X Using the model (test) known known unknown known known unknown
- 23. Trinity College Dublin, The University of Dublin 23 Supervised Learning Quote Slide Option 2 Classification example X (Data) Y (Labels – desired output) 1 1 1 1 2 2 2 2 3 3 3 3 New unseen image ML_model: y = f(X) ML model 2 Good prediction
- 24. Trinity College Dublin, The University of Dublin 24 Unsupervised Learning Quote Slide Option 2 Clustering example X (Data) ML model
- 25. Trinity College Dublin, The University of Dublin 25 What can we do with ML? Quote Slide Option 2 “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019 In the simplest case, the ML model learns just once. It will not learn later on or change automatically. So, it learns once and then it is used as an anomaly detection tool. X = [feature1, feature2, …, featurek]
- 26. Trinity College Dublin, The University of Dublin 26 What can we do with ML? Quote Slide Option 2 “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019 Reinforcement learning - Like humans, these algorithms learn over time (one step at a time) - Learn from the outcome (e.g., “good” or “bad”) of your actions. Learn from your successes and mistakes
- 27. Trinity College Dublin, The University of Dublin 27 What can we do with ML? Quote Slide Option 2 Real life data are messy! Noisy data, missing data, artifacts, mislabelled data. https://platerecognizer.com/blueiris/
- 28. Trinity College Dublin, The University of Dublin 28 Objectives and learning outcomes Quote Slide Option 2 • MLO1 Configure a programming environment suitable for exploring ML techniques • MLO2 Prepare datasets for ML processing, visualise the data, and understand the consequences of decisions made in cleaning data • MLO3 Assess the performance of a ML pipeline • MLO4 Critically evaluate the outputs of a ML pipeline • MLO5 Communicate with ML experts and non-experts: Explain goals and requirements of a project, interpret the outcomes of typical ML analyses, present results to non-experts. • MLO6 Assess the cost/benefit of distinct ML methodologies and explain what makes one approach more suitable than another one for a given task • MLO7. Understand challenges involving data sharing, storage, and privacy Problem/question Data collection Preprocessing / cleaning Analysing Interpretation / outcome Improve ML
- 29. Trinity College Dublin, The University of Dublin 29 What can we do with ML? Quote Slide Option 2 “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”, Aurélien Géron, 2019 Your background Groups