Fuzzy Logic and Fuzzy Systems Module Revision Fuzzy Logic and Fuzzy Systems Module Revision Chao Chen IMA1 and LUCID2 research groups
[email protected] 1Intelligent Modelling & Analysis 2Lab for Uncertainty in Data and Decisions Making 1 Fuzzy Logic and Fuzzy Systems Module Revision Overview Module Summary Topics Covered 2 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Summary Aims This module aims to provide a thorough understanding of fuzzy sets and systems from a theoretical and practical perspective. Objectives Introduce the theory and principles of fuzzy logic and fuzzy system Explain how fuzzy methods can be used to model uncertainty in real world examples Convey the properties and concepts underlying fuzzy inference systems and their applications Explore current research questions Provide practical experience on the design and implementation of fuzzy logic systems 3 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Basic Concepts Fuzzy sets: definitions, notation, membership functions... Properties of fuzzy sets: α-cuts, support, normality... Basic operations: complement, intersection, union... Parameterised operations: t-norm, t-conorm, operator families... 4 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Linguistic Variables Informal and formal definitions... Hedge: concentration (e.g. very), dilation (e.g. slightly)... Example membership functions (e.g. Triangular, Gaussian) Deriving terms, guidelines... Linguistic Truth: ture, false, unknown, undefined... Extension principle: allows a relationship from one domain to another 5 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Fuzzy Inference Systems (Mamdani) IF THEN Rules... Inference process: fuzzify inputs, combine inputs, implication, aggregate outputs, defuzzification Union and intersection operators: both for antecedents and antecedants Defuzzification: Numeric, Linguistic Common numeric deffuzification: Centre of Gravity, Mean of Maxima ... Problems with Defuzzification: other metrics 6 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Fuzzy Inference Systems (TSK) zeroth-order, first-order, higher order comparisons between Mamdani and TSK Tuning and Optimisation Fuzzy model identification: structure & parameters Evaluation (Selecting the ‘Best’): objective function, cost function, error measures... Search Methodologies: Exhaustive Search, Monte Carlo, Hill Climbing, Simulated Annealing... 7 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered ANFIS (adaptive network-based fuzzy inference system) model structure (essentially a TSK system) description of layers (calculations) parameters: antecedents, consequents the hybrid learning procedure (two passes) example models 8 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Type-2 Fuzzy Logic Type-2 Fuzzy Sets FootPrint of Uncertainty Interval Valued Fuzzy Sets Intersection, Union Type-Reduction, Defuzzification: algorithms, embedded sets... applications: supply chain modelling, decision making... 9 Fuzzy Logic and Fuzzy Systems Module Revision Module Overview Topics Covered Guest Lectures (not covered in the Exam) Realworld Applications (Prof. Jon Garibaldi) Similarity Measures on Fuzzy Sets and Applications (slides available on Moodle) Labs FuzzyR (need to have access for the exam) Generate/plot fuzzy sets/membership functions; build fuzzy models; 10 Fuzzy Logic and Fuzzy Systems Module Revision Exam Examinable Materials Materials provided over Moodle and Microsoft Teams What’s written on the slides What’s in the papers provided What’s in the videos What’s in the engagement activities What’s said/discussed/written/illustrated in the live sessions What’s in the lab tutorials What’s in the FuzzyR manual/help files What’s in the emails (e.g. announcement) 11 Fuzzy Logic and Fuzzy Systems Module Revision Exam Bloom’s Taxonomy The form of knowledge required to answer an exam question Knowledge: for what we often term ’bookwork’ based upon recall of factual information Comprehension: or ’understanding’, where students are asked to perform such actions as to describe, explain, classify ideas or concepts Application: for the situation where a student is asked to undertake an ’unseen’ task by using knowledge in a new way (or a variant upon one previously shown in lectures) 12 Fuzzy Logic and Fuzzy Systems Module Revision Exam Warnings/Tips Wherever you provide your reasoning or explanation, do not return lengthy answers. providing a lengthy answer may increase the chances that you make mistakes every erroneous comment in a solution/answer will get penalised Do not simply copy and paste. Answer each of the FOUR questions by starting a new page. We will be strict in the application of the marking criteria for both coursework/exams. 13 COMP4033-E1 COMP4033-E1 The University of Nottingham SCHOOL OF Computer Science A LEVEL 4 MODULE, SPRING SEMESTER 2020-2021 Fuzzy Logic and Fuzzy Systems Deadline for submission: see the Moodle submission box Open-book examination. Answer All FOUR Questions. Suggested time to complete the paper ~2 hours. This open-book examination will be marked out of 100. You may write/draw by hand your answers on paper and then scan them to a PDF file, or you may type/draw your answers into electronic form directly and generate a PDF file. Guidance on scanning can be found through the Faculty of Science Moodle Page Guidance for Remote Learning. Your solutions should include complete explanations and should be based on the material covered in the module. Make sure your PDF file is easily readable and does not require magnification. Make sure that each page is in the correct orientation. Text/drawing which is not in focus or is not legible for any other reason will be ignored. Submit your answers containing all the work you wish to have marked as a single PDF file, with each page in the correct orientation, to the appropriate dropbox on the module’s Moodle page. Use the standard naming convention for your document: [Student ID] [Module Code] [Academic Year]. Write your student ID number at the top of each page of your answers. Although you may use any notes or resources you wish to help you complete this open-book examination, the academic misconduct policies that apply to your coursework also apply here. You must be careful to avoid plagiarism, collusion or false authorship. Please familiarise yourself with the Guidance on Academic Integrity in Alternative Assessments, which is available on the Faculty of Science Moodle Page Guidance for Remote Learning. The penalties for academic misconduct are severe. Staff are not permitted to answer assessment or teaching queries during the period in which your examination is live. If you spot what you think may be an error on the exam paper, note this in your submission but answer the question as written.
欢迎咨询51作业君
