Künstliche Intelligenz (Blockveranstaltung)

Dozenten: Prof Dr Christoph Benzmüller, Prof Dr Christoph Schommer

Agenda, Part I

Monday, 1 April - Friday, 5 April

• Suchverfahren für die Lösung kombinatorischer Aufgaben
• Klassische und Nichtklassische Logiken und ihre Mechanisierung; DPLL, Resolution, Tableauxverfahren und Theorembeweisen
• Wissensbasierte- und Expertensysteme
• Mensch-Maschinen-Schnittstellen
• Mustererkennung insbesondere für interaktive Anwendungen
   

Agenda, Part II

Monday, 15 July (3 units)

• 09h15 (= cum tempore) - 12h15; Takustraße 9, Seminar Room 006.
• Content: Introduction and Course Overview / Part II - see Slides under "Resources": Introduction to Artificial Intelligence: is Eliza intelligent?, Shrdlu, sketching briefly different fields of Artificial Intelligence (weak vs strong AI; a word about :Reasoning, :Data Mining, :Natural Language Processing/Understanding, :Knowledge Representation, :Neural-based architectures), the ethical point, a need for an explainability, how robots change our life (short videos); What is CLAIRE?, Course Overview; Projects for Friday /
• Projects for Friday / Registration needed. Submission DEADLINE (presentation, extemded abstract): 19 July, 09h14.
   
• Afternoon ( 14h00 - 16h00 ): SAT Solver Competition (Prof Benzmueller); Takustraße 9, Seminarraum 006.

Tuesday, 16 July (4 units)

• 09h15 (= cum tempore) - 13h15; Takustraße 9, Seminar Room 006.
• Weak AI as the intersection between the field of Artificial Intelligence and the field of Data Science
• Example: Knowledge Discovery/Data Science
  • Data Science side: data preparation (e.g., data cleaning, data reduction, discretization), data understanding (e.g., visualization, use of statistics), data privacy (e.g., k-anonymity, l-diversity).
  • Artificial Intelligence side: Application of Machine Learning-algorithms.
    • Association Discovery and Market Basket Analysis (+ software demonstration)
    • Demographic Clustering and Profiling (+ software demonstration)
    • Bivariate Statistics

Wednesday, 17 July (3 units)

• 09h15 (= cum tempore) - 12h15; Takustraße 9, Seminar Room 006.
• Strong AI as the desire to transfer cognitive (human) intelligence to machines
• Example: Cognitive Simulation, Neural Architectures, Selected Aspects of Natural Language Processing - see Slides under "Resources"
  • Similarity measures: n-grams and Jaccard, Edit distance, Soundex
  • Text preparation: Stemming (Porter Stemmer), Tokenization, Part-of-Speech Tagging, Context-free Parsing.
  • Simulation of a human cognition: disambiguation of word senses by a connectionist architecture: competitive learning, spreading activation and lateral inhibition, sub-symbolic approach with symbolic values, interplay of several layers (syntax, lexical, semantic, context); realization of an associative mind-map towards an associative memory.

Thursday, 18 July (3 units)

• 09h15 (= cum tempore) - 11h30 (no break); Takustraße 9, Seminar Room 006.
• Individual discussion // Extended Abstract and the slides (or equivalent)
• With regard to the workshop, we have discussed several aspects in view of the content as well as organisational issues:
  • a. If you include videos in your presentation, you should include them in a reasonable frame of mind. Videos should be supportive, not but replace your lecture.
  • b. You are welcome to raise questions to the audience, if these support a better understanding, encourage the participants to think and/or meaningfully support the content.
  • c. If your presentation becomes multi-disciplinary (and not exclusively technical), please say so at the beginning.
  • d. Topics within a block (here: autonomous weapons, self-driving cars, AI in school and for an Education) should be coordinated as much as possible. Particularly, redundancies (the same picture, video,...) should - if possible - be avoided.
  • e. A suggestion (but is not binding)  regarding the Extended Abstract: introduce your topic in 2-3 lines (= meaning of the topic). Then, continue then with your hypothesis, and then reason the arguments. Finally, conclude with a short summary.
  • f. References in the text should be placed (in brackets, e.g. like this: [1] ) In the final booklet, I will then also publish your references directly under your text ( so that the assignment remains unique).
  • g. Since we have sent invitations to external parties, your slots remain as specified.
  • h. Regarding evaluation: it is important to me that your presentation follows the presented topic, has a clear structure and is able to become a general understanding.

Friday, 19 July (7 units)

• 09h15 (= cum tempore) - 14h15 ; Takustraße 9, Seminar Room 006.
• Public Workshop; individual presentations:
  • 15 minutes each
  • up to 10 minutes for Q&A
  • Language can be chosen: Deutsch oder Englisch
  • Submission Deadline (hard): 09h14, 19 July 2019
• Presentations
  • Almohamad Basel:
    Artificial Intelligence and self-driving cars (09h15)
  • Guillermo Vallejos Arana:
    Werden autonome Autos in Zukunft existieren? (09h45)
  • Eiad Rostom:
    Artificial Intelligence and Autonomous Weapons (10h15)
  • Dominik Blöse:
    Risks and Challenges of Autonomous Weapons (10h45)
  • Ulrike Bath:
    Auswirkungen der Künstlichen Intelligenz auf die Arbeitswelt/Arbeitsmarkt (11h15)
  • Khank Bui Trong:
    Artificial Intelligence in Architecture and Civil Engineering (11h45)
  • Dastan Kasmamytov:
    Künstliche Intelligenz im Gesundheitswesen (12h15)
  • Carmen Johansmeier:
    Can we become friends with robots? (12h45)
  • Dhruti Sawant:
    Artificial Intelligence and Education: how students will learn? (13h15)
  • Marvin Kleinert:
    Künstliche Intelligenz in der Schule (13h45)

Literatur

• Some parts of the course were underpinned by slides, whcih can be found under Resources.
• Other parts of the course were performed at the black board (see public literature).

Zusätzliche Informationen

• Evaluation Part 1: written or oral examination (50%)
• Evaluation Part 2: Presentation (30%; incl quality of slides) and Extended Abstract (20%) of up to 1200 words.

Voraussetzungen

• Grundkenntnisse in Mathematik und Algorithmen & Datenstrukturen.
• Grundsätzliches Interesse im Gebiet der Künstlichen Intelligenz