INF3490/INF4490: Biologically Inspired Computing – Autumn 2017 • Lecturer: – Kai Olav Ellefsen ( kaiolae@ifi.uio.no ) INF3490/INF4490 – Weria Khaksar ( weriak@ifi.uio.no ) Biologically Inspired Computing – Jim Tørresen ( jimtoer@ifi.uio.no ) Lecture 1 – 2017 Course Introduction • Lecture time: Monday 10.15-12.00 • Lecture room: OJD Simula Jim Tørresen • Group Lecture (starting this week): – Group 2: Wednesday 10:15-12:00 (OJD 1454 Computer Room Sed ) – Group 3: Thursday 10:15-12:00 (OJD 3418 Computer Room Limbo) – Group 1: Friday 10:15-12:00 (OJD 2443 Computer Room Modula) 2 • Course web page: www.uio.no/studier/emner/matnat/ifi/INF3490 INF3490/INF4490 Group Teachers Syllabus: • Selected parts of the following books (details on course Edvard Per Antoine Bjørn Ingeberg T or Jan Derek web page): Bakken Carlsen Fesche Berstad – A.E. Eiben and J.E. Smith: Introduction to Evolutionary Computing, Wednesday Thursday Friday Misc Second Edition (ISBN 978-3-662-44873-1). Springer. – S. Marsland: Machine learning: An Algorithmic Perspective, Second Edition, ISBN: 978-1466583283 – On-line papers (on the course web page). • The lecture notes. Obligatory Exercises: • Two exercises: Evolutionary algorithms (deadline 25 Sept) and Machine learning (deadline 20 Oct). • Announced on the course web page (Messages) two w eeks before the deadline. • Supervision: Group lectures and Slack (register at http://inf3490.slack.com using UiO e-mail address) 3 • Students registered for INF4490 will be given additional tasks in the two 4 exercises. This is the only difference compared to INF3490. 1
Lecture Plan Autumn 2017 (tentative) Supporting Literature in Norwegian (not syllabus) Date Topic Syllabus Jim Tørresen: hva er KUNSTIG INTELLIGENS 28.08.2017 Intro to the course. Optimization and search. Marsland (chapter 9.1, 9.4-9.6) Universitetsforlaget Nov 2013, ISBN: 9788215020211 04.09.2017 Evolutionary algorithms I: Introduction and representation. Eiben & Smith (chapter 1-4, not 1.4, 3.6 and 4.4.2) T opics: 11.09.2017 Evolutionary algorithms II: Population management and Eiben & Smith (chapter 5-6, not 5.2.6, 5.5.7, 6.5- popular algorithms 6.6 and 6.8) (+ Marsland 10.1-10.4) • Kunstig intelligens og 18.09.2017 Evolutionary algorithms III: Multi-objective optimization. Eiben & Smith (chapter 9, 10, 12, not 10.4 and Hybrid algorithms. Working with evolutionary algorithms. 12.3.4) intelligente systemer 25.09.2017 Intro to machine learning and classification. Single-layer Marsland (chapter 1 and 3, not 3.4.1) neural networks. • Problemløsning med kunstig 02.10.2017 Multi-layer neural networks. Backpropagation and practical Marsland (chapter 2.2 and 4) issues. intelligens 09.10.2017 Reinforcement learning and Deep Learning Marsland (chapter 11) + online paper • Evolusjon, utvikling og læring 16.10.2017 Support vector machines. Ensemble learning. Dimensionality Marsland (chapter 8, 13, 6.2.) reduction. • Sansing og oppfatning 23.10.2017 Unsupervised learning. K-means. Self-organizing maps. Marsland (chapter 14) • Bevegelse og robotikk 30.10.2017 Swarm Intelligence. Fuzzy logic. TBA (On-line papers on the course web page) • Hvor intelligente kan og bør 06.11.2017 Bio-inspired computing for robots and music. Future On-line papers on the course web page perspectives on Artificial Intelligence including ethical issues maskiner bli? 13.11.2017 Summary and Questions 5 6 What is the Course about? Self learning/Machine learning (ex: evolutionary computation) • Artificial Intelligence/Machine learning/Self-learning: – Technology that can adapt by learning Algorithm • Systems that can sense, reason (think) and/or respond • Inspired from biology/nature System to be • Increase intelligence in both single node and designed multiple node systems Data set/ Learning by specification examples 7 2
Data Driven Modeling in Machine “Future work” – Current ML/AI challenges Learning • Scalability • Development of general intelligent systems (larger range of problems) • Predictable behavior in unfamiliar situations • Battery life in portable products • Mechanical solutions for robots (soft material) 9 10 Man/Woman vs Machine – Who are smartest? Major Mechanisms in Nature • Evolution: Biological systems • Machines are good at: develop and change during – number crunching generations. – storing data and searching in data • Development/growth: By cell – specific tasks (e.g. control systems in division a multi-cellular organism manufacturing) is developed. • Humans are good at: • Learning: Individuals undergo – sensing (see, hear, smell etc and be able to learning through their lifetime. recognize what we senses) • Collective behavior: Immune – general thinking/reasoning systems, flocks of birds, fishes etc – motion control (speaking, walking etc). • Sensing and motion 11 • 3
What Methods are best? Artificial Intelligence Application Examples • Computer systems – Web search – Web shopping – Optimization e.g. the design of physical shapes – Route planning • Embedded/physical systems size/complexity – Smartphone user adaptation Increasing – Detecting faces/people smiling in cameras – Service robots – Driverless drones, cars and submarines 13 14 Google Driverless Car 15 16 4
Google Driverless Car (Inter) Active Music Indirect Control Direct Control o Use on-body sensors to adapt the o Navigate within the song music to the mood of the user o Control certain instruments (e.g. o Listen to music that pushes you to keep playing the chorus drumbeat work out harder in the verse) o Fuse the musical preferences of o Change the tempo of the song multiple users into one song 17 18 Apple app: https://itunes.apple.com/us/app/pherom usic/id910100415?ls=1& mt=8 Ant Colony Optimization (ACO) • Ants find shortest path to food source from nest. • Ants deposit pheromone along traveled path which is used by other ants to follow the trail. • This kind of indirect communication via the local environment is called stigmergy. 19 20 5
MECS: Multi-sensor Elderly Care Systems EPEC: Prediction and Coordination for 1 PhD (Trenton Schulz) + 2 postdocs (Weria Khaksar and Zia Uddin) Robots and Interactive Music (2015-2019) 1 PhD (Tønnes Nygaard) + 2 post-docs (Charles Martin Goal: Create and evaluate multimodal mobile human supportive and Kai Olav Ellefsen) 2015-2019 systems that are able to sense, learn and predict future events . Project consortium: • Robotics and Intelligent Systems group (coordinator) • DESIGN group (IFI) • National: Oslo Municipality (Oslo kommune, Gamle Oslo) o Norwegian Centre for Integrated Care and o Telemedicine (Tromsø) XCENTER AS (3D sensor) o Novelda AS (ultra wideband sensor) o • International: Goal: Design, implement and evaluate multi-sensor systems that University of Hertfordshire o are able to sense, learn and predict future actions and events. University of Reading Whiteknights o http://www.mn.uio.no/ifi/english/research/projects/epec http://www.mn.uio.no/ifi/forskning/prosjekter/mecs Funding: FRIPRO, Research Funding: IKTPLUSS, Council of Norway Research Council of Norway Is Terminator Coming Close? Repetiton Questions • What is machine learning? • Give some examples of intelligent mechanisms in nature 23 24 6
Recommend
More recommend
