Page 275 - AI Computer 10

u Orange Data Mining: Orange Data Mining is an open-source data visualisation, machine learning, and data
mining toolkit that enables users to visualise data and perform data mining and machine learning operations.

ORANGE DATA MINING TOOL

Orange Data Mining is an open-source data visualisation and analysis tool that provides a powerful yet user-
friendly environment for data mining and machine learning tasks. It provides a graphical user interface (GUI) that
enables users to interactively create data analysis workflows by utilising various components known as Widgets.
Each widget is designed to fulfill a specific role in the data analysis process.

Widgets in Orange Data Mining Tool

Some of the important and most commonly used widgets of Orange Data Mining are:
u Data Loading Widgets: Data Loading widgets are essential for importing datasets from various sources into
the system for analysis and modeling. These widgets, such as File, SQL Table, and Google Sheets, allow users
from various backgrounds to work with data effectively.
• File: This widget allows you to load data from files in various formats such as CSV, Excel, and SQL.

• Data Table: This widget displays loaded data in a tabular format.
u Data Exploration Widgets: Data Exploration widgets are vital tools that allow users to visualise, analyse, and
gain insights into their datasets. These widgets help users understand the underlying patterns, distributions,
and relationships within the data before proceeding to modeling.
• Scatter Plot: This widget helps to visualise the relationship between two variables.
• Data Table: This widget allows examination of individual records and helps in identifying missing values
or anomalies.

• Distributions: This widget displays histograms and other statistical distributions of variables.
u Preprocessing Widgets: Preprocessing widgets prepares raw data for analysis and modeling. These widgets
help you clean up the data, like filling in missing values or making sure all your data is on the same scale.
• Impute: This widget addresses missing values in the dataset.
• Normalize: This widget normalises the data to a common scale.
• Select Columns: This widget allows you to select specific columns from the dataset.

u Feature Selection Widgets: Feature Selection widgets help to identify and select the most relevant features
for modeling.

• Select Columns: This widget helps to select specific features manually based on domain knowledge or
exploratory analysis.

• Select Best Features: This widget helps to select the best features automatically based on certain criteria
like mutual information or correlation.

141
141