Elements of AI: Künstliche Intelligenz demystifiziert

Was sollte man als Laie über KI wissen? Der finnische MOOC „Elements of AI“ setzt sich zum Ziel das Thema der künstlichen Intelligenz zu demystifizieren und ein paar Grundlagen zu vermitteln.
Der Kurs kommt sehr gut an. Auch ich habe ihn gemacht, sogar bis zum Ende.

Hier ein paar Eckdaten zum MOOC:

MOOC: Elements of AI: https://www.elementsofai.com/
• entwickelt von der University of Helsinki und der Firma Reaktor
• Ziel: Verständnis für KI entwickeln: Verständliche Erklärungen mit Beispielen und Visualisierungen und Aufgaben
• > 230.000 TN
• Ziel: frei verfügbar in allen europäischen Sprachen; Übersetzungen kommen von der EU-Kommission
• 6 Kapitel, 25 Aufgaben

Hier ein Überblick der Inhalte

Kapitel 1 What is AI?

Lernziele
• Explain autonomy and adaptivity as key concepts for explaining AI
• Distinguish between realistic and unrealistic AI (science fiction vs. real life)
• Express the basic philosophical problems related to AI including the implications of the Turing test and Chinese room thought experiment

1) Definitionen / Anwendungsbeispiele

Inhalte:
Beispiele: Selbst fahrende Autos, Inhaltsempfehlungen, Bild- und Videoverarbeitung
Charakteristische Eigenschaften: Autonomie und Adaptivität

Aufgabe 1:  Is this AI or not?
Frage:
Is this AI? Which of the following are AI and which are not. Choose yes, no, or “kind of” where kind of means that it both can be or can’t be, depending on the viewpoint.

2) Verwandte Gebiete

Inhalte:
Machine learning, deep learning, data science, robotics

Aufgabe 2: KI-Taxonomie anhand eines Euler-Diagramms konstruieren
Frage:
Construct a taxonomy in the Euler diagram example given below showing the relationships between the following things: AI, machine learning, computer science, data science, and deep learning.

Aufgabe 3: Zuordnung von Beispielen zu Gebieten
Frage:
Consider the following example tasks. Try to determine which AI-related fields are involved in them. Select all that apply.

3) Philosophie

Inhalte:
Turing Test, Chinesisches Zimmer; Schwache / starke KI

Aufgabe 4: Eigene Definition von KI bzw. Bewertung von Definitionen
Frage:
Which definition of AI do you like best? How would you define AI?

Kapitel 2 AI problem solving

Lernziele
• Formulate a real-world problem as a search problem
• Formulate a simple game (such as tic-tac-toe) as a game tree
• Use the minimax principle to find optimal moves in a limited-size game tree

1) Suche und Probleme lösen

Inhalte:
Search in practice: getting from A to B; Spielzeugproblem: Chicken crossing, Türme von Hanoi

Aufgabe 5: A smaller rowboat
Frage:
Using the diagram with the possible states below as a starting point, draw the possible transitions in it; Having drawn the state transition diagram, find the shortest path from NNNN to FFFF, and calculate the number of transitions on it.

Aufgabe 6: Die Türme von Hanoi
Frage:
Draw the state diagram.

2) Probleme mit AI lösen

Inhalte:
Geschichtlicher Rückblick

3) Suche und Spiele

Inhalte:
Beispiel tic tac toe; Game trees; Minimax Algorithmus;

Aufgabe 7: tic tac toe
Frage:
Enter the value of the game as your answer.

Kapitel 3 Real world AI

Lernziele
• Express probabilities in terms of natural frequencies
• Apply the Bayes rule to infer risks in simple scenarios
• Explain the base-rate fallacy and avoid it by applying Bayesian reasoning

1) Odds and probability

Inhalte:
Probability, odds

Aufgabe 8: probabilistische Vorhersagen
Frage:
Consider the following four probabilistic forecasts and outcomes. What can we conclude based on the outcome about the correctness of the forecasts? 

Aufgabe 9: Odds
Frage:
For the first three items 1–3, convert from odds to probabilities expressed as natural frequencies; for example from 1:1 to 1/2. Give your answer as a fraction, for example 2/3.
For the last three items 4–6, convert the odds into probabilities expressed as percentages (e.g. 4.2%). Give your answer in percentages using a single decimal, for example 12.2%.

2) The Bayes rule

Inhalte:
Prior and posterior odds, likelihood ratio

Aufgabe 10: Bayes rule (part 1 of 2)
Frage:
Apply the Bayes rule to calculate the posterior odds for rain having observed clouds in the morning in Helsinki.

Aufgabe 11: Bayes rule (part 1 of 2)
Frage:
Calculate the posterior odds for her having breast cancer using the Bayes rule. 

3) Naive Bayes classification

Inhalte:
Spam filters

Aufgabe 12: One word spam filter
Frage:
Calculate the posterior odds for spam given this word using the table above, starting from prior odds 1:1.

Aufgabe 13: Full spam filter
Frage:
Now use the naive Bayes method to calculate the posterior odds for spam given the message “million dollars adclick conferences”.

Kapitel 4 Machine learning
Lernziele
• Explain why machine learning techniques are used
• Distinguish between unsupervised and supervised machine learning scenarios
• Explain the principles of three supervised classification methods: the nearest neighbor method, linear regression, and logistic regression

1) The types of machine learning

Inhalte:
Klassisches Beispiel Handschriftenerkennung; Typen des Machine learning

2) Nearest neighbor classifier

Aufgabe 14: Customers who bought similar products
Frage:
Calculate the similarity of Travis relative to the six users in the training data, identify the user who is most similar to Travis by selecting the largest of the calculated similarities. Predict what Travis is likely purchase next by looking at the most recent purchase

Aufgabe 15: Filter bubbles
Frage:
Do you think that filter bubbles are harmful? After all, they are created by recommending content that the user likes. What negative consequences, if any, may be associated with filter bubbles?
Think of ways to avoid filter bubbles while still being able to recommend content to suit personal preferences. Come up with at least one suggestion. 

3) Regression

Inhalte:
linear / logistic regression

Aufgabe 16: Linear regression
Frage:
Calculate the life expectancies for the following example cases

Aufgabe 17: Life expectancy and education (part 1 of 2)
Frage:
Given the data, what can you tell about the life expectancy of people who have 15 years of education?

Aufgabe 18: Life expectancy and education (part 2 of 2)
Frage:
Which of the following options would best match your estimate of the life expectancy for people with 15 years of education? Choose the most specific option that you think is justified by fitting the straight line model to the above data.

Aufgabe 19: Logistic regression
Frage:
If you wanted to have an 80% chance of passing a university exam, based on the above figure, how many hours should you approximately study for?

Kapitel 5 Neural networks

Lernziele
• Explain what a neural network is and where they are being successfully used
• Understand the technical methods that underpin neural networks

1) Neural network basics

Inhalte:
Elements of a neural network, deep learning

Aufgabe 20: Elements of a neural network
Frage:
Label the different components of a neuron into the diagram below

2) How neural networks are built

Aufgabe 21: Weights and inputs
Frage:
In this exercise, consider the following expression that has both weights and inputs: 10.0 + 5.4 × 8 + (-10.2) × 5 + (-0.1) × 22 + 101.4 × (-5) + 0.0 × 2 + 12.0 × (-3) = -543.0

Aufgabe 22: Activations and outputs
Which of the activations described above gives:
the largest output for an input of 5?
the smallest output for an input of -5?
the largest output for an input of -2.5?

3) Advanced neural network techniques

Inhalte:
Convolutional neural networks, generative adversarial networks.

Kapitel 6 Implications

Lernziele
• Understand the difficulty in predicting the future and be able to better evaluate the claims made about AI
• Identify some of the major societal implications of AI including algorithmic bias, AI-generated content, privacy, and work

1) About predicting the future

Aufgabe 23: What is the perception of AI?
Frage:
For this exercise, we want you to think about how AI is portrayed. Do an online image search for the term “AI” and see what kinds of pictures come up. If you are using Google search, you should choose „Images“ in the top of the screen.
What’s the general impression you get about AI from the image search results? Is this an accurate representation of AI? Why or why not?

2) The societal implications of AI

Inhalte:
Algorithmic bias, Seeing is believing, Privacy, Work

Aufgabe 24: Implications of AI
Frage:
What kind of articles (in newspapers and magazines or other popular science outlets such as blogs, …) are being written about AI – and do you think they are realistic? Do an online search about AI related to one of your interests. Choose one of the articles and analyze it.
Mention the title of the article along with its author and where it was published (as a URL if applicable) in your answer.
Explain the central idea in the article in your own words using about a paragraph of text (multiple sentences.)
Based on your understanding, how accurate are the AI-related statements in the article? Explain your answer. Are the implications (if any) realistic? Explain why or why not.

3) Summary

Aufgabe 25: AI in your life
Frage:
How do you see AI affecting you in the future, both at work and in everyday life? Include both the positive and possible negative implications.

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.

* Die DSGVO-Checkbox ist ein Pflichtfeld

*

Ich stimme zu.