Page 144 - AI Computer 10
P. 144

There are datasets readily available from various research institutions, government agencies, or online
              platforms that can be used for common AI tasks like image classification, natural language processing,
              etc.


                  Pop
                  Pop Quiz                                     Quiz
              Write at least five data sources to acquire data for an AI project.

              1.  ____________________________________                 2.   ____________________________________

              3.  ____________________________________                 4.   ____________________________________

              5.  ____________________________________

            DATA EXPLORATION

            Data Exploration  is  the third  stage of  the AI Project Cycle where collected data needs to  be explored and
            interpreted to extract useful information and better understand the quality and nature of data. Let us understand
            the concept of data exploration with the help of an example.
            Example: Suppose you are the head of the Admission Cell in a school. All the admission counselors have submitted
            their data in the form of Excel sheets. In such a case, you get a large amount of data among which both useful
            and useless data is available. Naturally, you want to keep only the useful data. In such a situation, you need to
            spend time, exploring the data to extract useful data according to your requirements. Similarly, when you have
            a large number of datasets about a particular topic, you can do the following things:

             u Identify trends, relationships, and patterns present in the dataset.
             u Define an efficient strategy to extract the relevant data from the dataset.

            Tools for Data Visualisation

            Data Visualisation can be thought of as magic wands that turn dull numbers and data into vibrant pictures. These
            tools act as artists, making data easy to understand by creating visual masterpieces. Think of them as our trusty
            companions in the journey of AI, revealing patterns and stories hidden in the numbers. Let’s explore how these
            tools make the world of data not just informative, but also fascinating and fun! Some popular and free data
            visualisation tools are:
             u Microsoft  Excel: Excel,  a spreadsheet  application  from Microsoft, is a manual  data exploration  tool.  It
                provides different types of charts and objects for visualising data. Although, it is not preferable for big
                datasets, it can still provide insights into small to medium datasets.
             u Microsoft Power BI: Microsoft Power BI is a data visualisation tool offered by Microsoft. It is a business
                intelligence (BI) tool that helps you analyse, visualise, and share data. It can work with datasets of any size,
                clean and transform the data, create charts, graphs and other visuals for the data, and also allows to share
                insights to help with decision-making. It is freely available to download and use.
             u Looker Studio: Looker Studio, formerly known as Google Data Studio, is a free cloud-based tool that helps
                users create reports and dashboards from data. It can connect up to 12 data sources simultaneously and
                provides widgets, charts, graphs and maps to visualise data in different forms.
             u Datawrapper: Datawrapper is a data visualisation tool that lets users create interactive charts, maps, and
                tables without any coding. It’s designed for journalists, researchers, businesses, and educators, making it
                easy to turn raw data into visually appealing graphics, in minutes.

                10
                10
   139   140   141   142   143   144   145   146   147   148   149