Book Image

Machine Learning with Qlik Sense

By : Hannu Ranta
Book Image

Machine Learning with Qlik Sense

By: Hannu Ranta

Overview of this book

The ability to forecast future trends through data prediction, coupled with the integration of ML and AI, has become indispensable to global enterprises. Qlik, with its extensive machine learning capabilities, stands out as a leading analytics platform enabling businesses to achieve exhaustive comprehension of their data. This book helps you maximize these capabilities by using hands-on illustrations to improve your ability to make data-driven decisions. You’ll begin by cultivating an understanding of machine learning concepts and algorithms, and build a foundation that paves the way for subsequent chapters. The book then helps you navigate through the process of framing machine learning challenges and validating model performance. Through the lens of Qlik Sense, you'll explore data preprocessing and analysis techniques, as well as find out how to translate these techniques into pragmatic machine learning solutions. The concluding chapters will help you get to grips with advanced data visualization methods to facilitate a clearer presentation of findings, complemented by an array of real-world instances to bolster your skillset. By the end of this book, you’ll have mastered the art of machine learning using Qlik tools and be able to take your data analytics journey to new heights.
Table of Contents (17 chapters)
1
Part 1:Concepts of Machine Learning
6
Part 2: Machine learning algorithms and models with Qlik
12
Part 3: Case studies and best practices

Introduction to Qlik tools

Qlik Sense is a leading data analytics and business intelligence platform and contains many tools and features for data analytics relating to machine learning. In this chapter, we will take a closer look at the key features of the Qlik platform.

Machine learning and AI capabilities on the Qlik platform can be divided into three different components:

  • Insight Advisor
  • Qlik AutoML
  • Advanced Analytics Integration

Insight Advisor

Qlik Insight Advisor is a feature of Qlik Sense that uses natural language processing (NLP) and machine learning to help users explore and analyze data more effectively. It allows users to ask questions about their data in natural language and to receive insights and recommendations in real time. It also auto-generates advanced analytics and visualizations and assists with analytics creation and data preparation.

Insight Advisor utilizes a combination of Qlik’s associative engine and augmented intelligence engine and supports a wide range of use cases, as seen in the following figure:

Figure 1.1: Qlik Insight Advisor and different scenarios

Figure 1.1: Qlik Insight Advisor and different scenarios

Did you know?

The Qlik associative engine is the core technology that powers the Qlik data analytics and business intelligence platform. It is a powerful in-memory engine that uses an associative data model, which allows users to explore data in a way that is more intuitive and natural than traditional query-based tools.

Instead of pre-defined queries or data models, the engine automatically associates data across multiple tables and data sources based on common fields or attributes and uses a patented indexing technology that stores all the data in memory, enabling real-time analysis and exploration of even the largest datasets. It is a powerful and innovative technology that underpins the entire Qlik platform.

Insight Advisor has the following key features:

  • Advanced insight generation: Insight Advisor provides a way to surface new and hidden insights. It uses AI-generated analyses that are delivered in multiple forms. Users can select from a full range of analysis types, which are auto-generated. These types include visualizations, narrative insights, and entire dashboards. Advanced analytics is also supported, and Insight Advisor can generate comparison, ranking, trending, clustering, geographical analysis, time series forecasts, and more.
  • Search-based visual discovery: Insight Advisor auto-generates the most relevant and impactful visualizations for the users, based on natural language queries. It provides a set of charts that users can edit and fine-tune before adding to the dashboard. It is context-aware and reflects the selections with generated visualizations. It also suggests the most significant data relationships to explore further.
  • Conversational analytics: Conversational analytics in Insight Advisor allows users to interact using natural language. Insight Advisor Chat offers a fully conversational analytics experience for the entire Qlik platform. It understands user intent and delivers additional insights for deeper understanding.
  • Accelerated creation and data preparation: Accelerated creation and data preparation helps users to create analytics using a traditional build process. It gives recommendations about associations and relationships in data. It also gives chart suggestions and renders the best types of visualizations for each use case, which allows non-technical users to get the most out of the analyzed data. Part of the data preparation also involves an intelligent profiling that provides descriptive statistics about the data.

Note

A hands-on example with Insight Advisor can be found in Chapter 9, where you will be given a practical example of the most important functionalities in action.

Qlik AutoML

Qlik AutoML is an automated machine learning tool that makes AI-generated machine learning models and predictive analytics available for all users. It allows users to easily generate machine learning models, make predictions, and plan decisions using an intuitive, code-free user interface.

AutoML connects and profiles data, identifies key drivers in the dataset, and generates and refines models. It allows users to create future predictions and test what-if scenarios. Results are returned with prediction-influencer data (Shapley values) at the record level, which allows users to understand why predictions were made. This is critical to take the correct actions based on the outcome.

Predictive data can be published in Qlik Sense for further analysis and models can be integrated using Advanced Analytics Integration for real-time exploratory analysis.

Using AutoML is simple and does not require comprehensive data science skills. Users must first select the target field and then AutoML will run through various steps, as seen in the following figure:

Figure 1.2: The AutoML process flow

Figure 1.2: The AutoML process flow

With the model established and trained, AutoML lets users make predictions on current datasets. Deployed models can be used both from Qlik tools and other analytics tools. AutoML also provides a REST API to consume the deployed models.

Note

More information about AutoML, including hands-on examples, can be found in Chapter 8.

Advanced Analytics Integration

Advanced Analytics Integration is the ability to integrate advanced analytics and machine learning models directly into the Qlik data analytics platform. This integration allows users to combine the power of advanced analytics with the data exploration and visualization capabilities of Qlik to gain deeper insights from their data.

Advanced Analytics Integration is based on open APIs that provide direct, engine-level integration between Qlik’s Associative Engine and third-party data science tools. Data is being exchanged and calculations are made in real time as the user interacts with the software. Only relevant data is passed from the Associative Engine to third-party tools, based on user selections and context. The workflow is explained in the following figure:

Figure 1.3: Advanced analytics integration dataflow

Figure 1.3: Advanced analytics integration dataflow

Advanced analytics integration can be used with any external calculation engine, but native connectivity is provided for Amazon SageMaker, Amazon Comprehend, Azure ML, Data Robot, and custom models made with R and Python. Qlik AutoML can also utilize advanced analytics integration.

Note

More information, including practical examples about advanced analytics integration, can be found in Chapter 7. Installing the needed components for the on-premises environment is described in Chapter 5.