AI project cycle

AI project cycle

1. Problem Scoping

   • Data Acquisition
     • Data Exploration
       • Modelling
         • Evaluation
           • In this level you put random photos into the machine as it to find out what it is and display it.
             • The module giving the best data is the one you chose.
         • Modelling is made up of AI, ML and DL.
           • Ai is the umbrella of ML and DL.
             • AI can be both a Rule-based and a Learning-based Approaches.
               • At this level the cleansed data is put into an AI-enabled machine.
       • Here we try to find patterns and relationships between the data provided to us.
         • We can also define the strategy to be used for the later stage.
           • This is a very graphical representation. Graphs like histograms and bar graphs are used.
             • Here the data is cleansed to get the best results. This process is called Data Cleansing.
     • Needs to be acquired from authentic, reliable, correct and open-sourced websites.
       • Can be acquired from Sensors, APIs, Web Scraping, Surveys, Observations and Cameras.
         • Survey Maps: To represent data like a flowchart Survey Maps are used. Here everything is interconnected. 4 key things are used. They are Circles, Arrows, Loops and '+' and '-' signs.
   • Who (Tells who are the stakeholders of the problem)
   • What (Explains if the problem and if it is a problem)
   • Where (Explains the context of the problems)
   • Why (Explains how it will benefit the stakeholders)
   • A problem can be scoped with the help of Ishikawa diagram or the 4Ws canvas.