System-modeling

A Miau Wiki wikiből
A lap korábbi változatát látod, amilyen Pitlik (vitalap | szerkesztései) 2024. március 9., 08:10-kor történt szerkesztése után volt. (→‎General challenge for the 2024-spring-semester)

General challenge for the 2024-spring-semester

Optimized OAM-generator - context free 
- https://miau.my-x.hu/bprof/2024/tavasz/general_challenge (description)
- https://miau.my-x.hu/miau/309/Excel_Sudoku_Solver_korretura.docx (typical anomalies in project documentations)

Recommended learning materials for autodidactic learning

International literature

MIAU-objects

Avatar-based learning materials

FAQ concerning the course

  • Q8: Is it possible to work in team?
  • A8: YES! (But the thesis should always be a unique product of one single author.)
  • Q9: Is this course really lecture-less?
  • A9: Yes, we have to cooperate concerning real problems. The theoretical aspects will be integrated into the creation-processes.
  • Q10: What is the MOTTO of the BPROF-education?
  • A10: Learning by doing!
  • Q11: What is the most important keyword of the BPROF-education?
  • A11: Sovereignty = willingness for problem solving mostly alone and based on all sources but at first always through own interpretations
  • Q12: Is it possible to close a course sooner, than the official end of a semester?
  • A12: Yes!
  • Q13: ...
  • A13: ...

Personalized tasks/wiki-articles with system modeling challenges

RISK INDEX FOR HOME-OFFICE-WORKERS

SUPPORTING FARMERS

Online dietetic expert (ODE)

Background information: https://miau.my-x.hu/mediawiki/index.php/BPROF_Thesis1
  • How to model anti-discrimination? (Which cheese is better than the other one concerning dietetic benchmarks/expectations/recommendations?)
    • Demo: XLSX based on random data (asap)
      • Version#1: OAM with raw data (ingredients %) = prima primissima award for cheese-variations (c.f. cheese of the year)
      • Version#2: OAM with differences compared to the expectations coming from the dietetic expert
  • How to model price/performance ratios? (Which product is worth consuming?)
    • Demo: XLSX based on random data (asap)
      • Version#1: OAM with raw data (ingredients %) = prima primissima award for cheese-variations concerning economic aspects (c.f. cheese of the year) / Y = price
      • Version#2: OAM with differences compared to the expectations coming from the dietetic expert / Y = price
  • How to optimize/model animal/human feed/food ratios? (Which mixture best covers the expectations?)
    • Demo: XLSX based on random data (asap)
  • How to derive the top 'n' attribute with the highest impact e.g., for product price? (this challenge is part of the general problem about the price/performance-ratios)
    • Demo: XLSX based on random data (asap)
      • Version#1: OAM with raw data (ingredients %) = prima primissima award for cheese-variations concerning economic aspects (c.f. cheese of the year) / Y = price
      • Version#2: OAM with differences compared to the expectations coming from the dietetic expert / Y = price
  • ...
    • Demo: ...

A bad (benchmarking) example and a better (good?) example for clarification of model-structures

  • description of an innovative chat-bot-development in a massive journalistic way: https://miau.my-x.hu/miau/295/gpt_demo.docx
    • Conclusions:
      • it seems to be trivial, which inputs (given tokens) lead to what kind of outputs (expected token(s))
      • it is here and now still not given: what is the process between inputs and outputs
  • description of a service having a payment policy and demo-views: https://miau.my-x.hu/miau2009/index_tki.php3?_filterText0=*chat-bot
    • KNUTH-principle: https://miau.my-x.hu/miau2009/index_tki.php3?_filterText0=*knuth
      • if somebody says, I do understand something, it is necessary to be capable producing a demo e.g. at least with random data or based on flow-charts, formulas, etc.
    • Conclusions:
      • it does not seem to be trivial, which inputs (e.g., sentences of employees?) lead to what kind of outputs (expected answers/reaction? - but about what kind of content? why?)
      • it is here and now also still not given: what is the process between inputs and outputs
    • KNUTH-principle: https://miau.my-x.hu/miau2009/index_tki.php3?_filterText0=*knuth
      • if somebody says, I do understand something, it is necessary to be capable producing a demo e.g. at least with random data or based on flow-charts, formulas, etc.

Expert system for fine-tuning IT-security systems (ES-FITS)

Necessity of a middle-ware for banks (NMB)

Automation of Incident Response Planning in IT Security (AIR-P-ITS)

ADAPTATION CHALLENGE: https://miau.my-x.hu/bprof/2023/adaptation_challenge.docx
Adapted solution v1: https://miau.my-x.hu/bprof/2023/company_ideality_index.xlsx
Expertise (with formula=model): https://miau.my-x.hu/bprof/2023/company_ideality_index_formula.xlsx
Expertise (with formula=model): https://miau.my-x.hu/bprof/2023/company_ideality_index_formula (2).xlsx

Conclusions:

  • it is important to be capable of identifying OAMs (objects and attributes with real data - e.g., log-data) - (c.f. adaptation-challenge)
  • it is important to be capable of deriving data processing alternatives (depends on real goals)
  • it is important to be capable of automation of data collection and data processing (where algorithms are just planned in a proper way if chatGPT can also write codes)
  • ...

Antivirus project

Health care project

Log-analysis

LINUX-bubble?!

Detecting Vehicle Malfunctions using Acoustic Data and AI

NCM-project & Symmetry-project

Hydrogen-project

Robot-eye for kicker-talent-identification (REFKI)

Robot-citizen

...

  • ...

Integration of Google Trends time-series in different ways for self-check-mechanisms

  • https://miau.my-x.hu/miau/297/google_trends_data_integration_variations.xlsx
    • there is no transformation needed if the dominant object (SK) is the anchor for each CSV
    • if the anchor is not the dominant object (c.f. HU), then the necessary transformation may to lead to the same results
  • self-check-layers:
    • the dominant object have to have always the same data independent from the parallel available objects in the queries
    • the non-dominant-object-based calculation have to lead to the same results as in case of the dominant object (without transformation needs)
      • (where the sameness as such must be interpreted as a kind of approximation with potential +/-1 unit difference because of integer values of the CSV-files)
  • https://miau.my-x.hu/miau/287/tutorial_google_trends_5plus.xlsx
    • programming concepts
Remarks: system models can be defined in different complexity levels (c.f. optimized estimation of risk-potentials based on raw data vs. checking/preparation of raw data)

...

  • ...

Other courses

Annexes

Structure of MIAU-WIKI:

Useful WIKI-articles in general:

Parallel articles in MIAU-WIKI:

Recommended literature: 

Evaluation of projects: https://miau.my-x.hu/bprof/2023/lectureless_education_evaluation_sheet.xlsx

Advanced levels

CONCLUSIONS: https://miau.my-x.hu/miau/297/modelling_challenges_on_the_field_of_IT_security_based_on_AI.docx

Gamification phase

1. game

Simulated conversation (where you have to be capable of playing both the role of the Student and the role of the "Chatbot":


  • Student#2:
  • Please, help me to interpret your solution!
  • Why your solution can be seen as a parallel solution?
  • "Chatbot"#2:
  • You have defined a dataset and a solution (see your own question before).
  • I have calculated a new solution based on the same data (as you asked me) and for the same analytical goal (as you asked me).


  • Student#3:
    • What is the common goal of the modelling in my documents and in your solution?
  • "Chatbot"#3:
  • The common goal/question/problem is:
  • How to derive an OAM-oriented models (an estimation-sets)
  • for different phases of different bearnings (as objects)
  • based on different (in the basic documentations pre-definied) acoustic attributes
  • in order to calculate a time-value (c.f. forecasting value) needed for a bearing to achieve a given level of noises.
#1-2-3 can be seen as a kind of initializing step. From now on (#4-5-6-7-8-...), real Students can formulate own questions:
  • Student#4:
    • What do you mean about OAM-oriented models?
  • "Chatbot"#4:
    • OAM = object-attribute-matrix is the structural core of the modelling here and now.
    • In our case:
      • objects are the different bearnings and their life-periodes/intervals/units
      • attributes are different characteristics of the measured noises (Xi) and the time needed to achieve a given noise-level
    • There are two types of attributes:
    • X-attributes: so-called independent variables
    • Y-attribute(s): the time needed to achieve a given noise-level, as the dependent variable
    • The Xi-attributes must have directions: based on
      • the-more-the-more and/or
      • the more-the-less principle (c.f. ceteris paribus expectations).
    • Theoretically, each problem may be transformed into (different) OAMs! (c.f. GPS = general problem solving)


  • Student#5:
    • What means the expression of ceteri paribus?
  • "Chatbot"#5:
    • Ceteris paribus does mean Ceteris paribus or caeteris paribus ( New Latin: [ˈse.tɛ.ris ˈpa.ri.bus]) is a Latin phrase meaning "other things equal"; English translations of the phrase include " all other things being equal " or " other things held constant " or " all else unchanged ". Ceteris paribus - Wikipedia (en.wikipedia.org/wiki/Ceteris_paribus)
    • In case of system modelling challenges, the expression of ceteris paribus highlights that between the attributes (Xi and/or Y) the relationships should always be checked, controlled, monitored, ruled, influenced in order to be capable of understanding complex knowledge representations forms (like neural networks, staircase functions, fuzzy systems, etc.).


  • Student#6:
    • What is a staircase function?
  • "Chatbot"#6:
    • A staircase function is a more complex knowledge representation than a regression model where an attribute has only one single parameter ("a") (c.f. y=a*x+b) - it means instead of "a" in a regression function, a staircase function is using a function where the X axis is the set of the ranks of the well-known input-values concerning the observed attribute and the Y-axis is the parameter value for each ranking level where these parameter-values may be substituted through the same value in case of more input levels.


  • Student#7:
    • Why exactly these objects (bearings*time) are predefined?
  • "Chatbot"#7:
    • The object-building is mostly an arbitrary process. The modell developer has to ensure defnining as many objects as possible.
    • There are data about just a few bearings.
    • Therefore, the bearing's life-cycle shouls/can/must be divided into time-intervals in order to have more objects concerning one single bearing.



  • Student#10:
    • What is the common in the two models?
  • "Chatbot"#10:


  • Student#11:
  • "Chatbot"#11:


  • Student#12:
  • "Chatbot"#12:


  • Student#13:
  • "Chatbot"#13:


  • Student#14:
  • "Chatbot"#14:


  • Student#15:
  • "Chatbot"#15:


  • Student#16:
  • "Chatbot"#16:

2. game

Please, choose a new sub-chapter: https://miau.my-x.hu/mediawiki/index.php/System-modeling#Personalized_tasks.2Fwiki-articles_with_system_modeling_challenges
And please, create a new conversation!
A lap eredeti címe: „https://miau.my-x.hu/mediawiki/index.php?title=System-modeling&oldid=83531”