Page 271 - AI Ver 3.0 class 10_Flipbook
P. 271

Transform Widgets

                 Transform Widgets are used for modifying and transforming data in your machine learning workflow. They allows
                 users to apply various preprocessing and feature engineering techniques to manipulate datasets, preparing them
                 for advanced analysis or model building. In short, we can say that these widgets help perform different operations
                 on data.

                 Some of these Transform widgets are:
                    • Data Sampler  Widget:  It  is  used  to  sample  or  split  a  dataset  into  smaller  subsets.  It  is  useful  when  you
                   want to work with a random subset of your data for training or testing purposes, or when you want to perform
                   cross-validation.
                    • Select Columns  Widget:  It  is  used  to  select  specific  attributes  (columns)  from  a
                   dataset that you want to work with, while excluding others. This is helpful for feature
                   selection, when you want to include only the relevant features for training a model.
                    • Impute Widget: It is used to handle missing data in your dataset. Missing values are
                   a common problem in real-world datasets, and imputation is the process of replacing
                   these missing values with some form of estimate (e.g., mean, median, mode).
                    • Discretize Widget: It is used to convert continuous data (like numbers) into categories
                   (like groups or ranges), making it easier for certain machine learning models to process
                   and understand the data.

                 Visualize Widgets

                 Visualize Widgets create charts and graphs that allow you to visualise patterns, distributions, and relationships
                 in your data. Visualization is an essential part of the data exploration process, as it can reveal insights that are not
                 easily apparent from raw data alone.

                 Some of these Visualize widgets are:
                    • Tree Viewer Widget: It is used to visualize decision trees. A decision tree is a popular
                   type of machine learning model used for making predictions. It’s like a flowchart that
                   helps make decisions based on different rules or conditions.
                    • Box Plot Widget: It is used to show the distribution of a continuous numerical feature.
                   It displays the median, quartiles, and outliers, giving a clear picture of how the data is
                   spread.
                    • Scatter Plot Widget: It is used to visualize the relationship between two continuous
                   numerical features. This widget is ideal for detecting correlations, trends, and clustering
                   within the data.
                    • Bar Plot Widget: It is used to visualize the distribution of categorical data by displaying
                   the frequency or count of categories. It provides a clear and simple way to compare the
                   size of different categories.

                 Model Widgets

                 Model widgets are used to create, train, and evaluate machine learning models. These widgets are essential for
                 building and testing models that make predictions based on the input data.  For example, you might use a model
                 to predict whether someone will like a product, or how much a house might cost based on its features.







                                                                                      Statisical Data (Practical)  269
   266   267   268   269   270   271   272   273   274   275   276