Supercharge Your Data Insights with Machine Learning Services

Table of Contents

1. Introduction
2. Insight Advisor and Analysis Types
   • Creating Recommendations and Visualizations with Insight Advisor
   • Overview of Analysis Types
   • Guided Authoring Experience with Analysis Types
3. Machine Learning Services in Qlik Sense
   • Forecasting
   • Time Series Decomposition
   • Clustering
   • Anomaly Detection
4. Demo: Clustering Capability in Qlik Sense
5. Feature Release: Key Drivers
6. Future Developments: Proactive Insights
7. Conclusion
8. Resources

Introduction

Welcome to the fourth installment of our video series on AI and ML at Qlik Sense. In this video, we will explore how We Are leveraging machine learning services in our platform and discuss some upcoming releases. My name is Nasser Compass, and I am a product manager for augmented analytics at Qlik.

In the previous demo, we covered the analysis types in Insight Advisor, which creates recommendations and visualizations using the click engine and the application's logical model. In this article, we will Delve deeper into the topic, providing a comprehensive overview of machine learning services in Qlik Sense and how they enhance our analysis capabilities.

Insight Advisor and Analysis Types

Insight Advisor is a powerful tool that offers users multiple interactions for generating insights. It includes three main functionalities: Insight Advisor Chat, Insight Advisor Search, and Analysis Types. Analysis Types, a new offering from Qlik, provides users with a guided authoring experience, making it easier to Create charts and visualizations.

With Insight Advisor, even users who are not familiar with the intricacies of data analysis can leverage its advanced capabilities. For instance, let's say a user wants to create a breakdown Chart but is unsure how to proceed. Insight Advisor handles all the necessary steps, from selecting the Relevant measures and Dimensions to creating the chart. Users can directly work with the chart within the sheet and explore its details using Qlik Sense's powerful exploration capabilities.

Qlik Sense continues to expand its range of analysis types, with over 21 options currently available. What sets these analysis types apart is the integration of machine learning services, which enhance their predictive and decision-making capabilities.

Forecasting

One example of an analysis Type that utilizes machine learning services is forecasting. Qlik Sense supports multiple engine functions for calculating forecasts, but the automation of the creation process is achieved through an analysis type. As long as the data model contains historical time-Based data, future events can be predicted with accuracy.

Time Series Decomposition

Time series decomposition, also known as STL (Seasonal and Trend decomposition using Loess), is another analysis type that leverages machine learning services. This technique allows users to break down complex data sets or problems into manageable components. The four key components of time series decomposition are observed, trend, seasonality, and noise. By isolating these components, users can gain valuable insights into the underlying Patterns within the data.

Clustering

Clustering is a machine learning technique used in several analysis types. It involves grouping similar data points based on their properties or features. The rationale behind clustering is that data points within the same group should exhibit similar characteristics. By clustering data, users can uncover Hidden patterns and relationships, enabling them to make informed decisions.

Anomaly Detection

Anomaly detection is widely recognized as one of the most common use cases for machine learning. Its purpose is to identify deviations within a data set. Anomaly detection is frequently used in fraud detection, financial analysis, data quality assessment, and cybersecurity. By employing machine learning services, Qlik Sense allows users to efficiently identify anomalies and take appropriate actions.

These are just a few examples of how machine learning services are integrated into Qlik Sense's analysis types. The dynamic nature of our platform ensures that new and exciting releases are constantly being developed and integrated.

Demo: Clustering Capability in Qlik Sense

Let's now dive into a demonstration of the clustering capability within Qlik Sense. For this demo, we will open a ticket consumption application and navigate to Insight Advisor. Once inside Insight Advisor, we will explore the analysis types and select clustering with k-means.

Within the clustering analysis type, we will choose the appropriate number of measures and dimensions. In this case, we will select one dimension (Agent) and two measures (ticket age and the number of tickets). After making these selections, the results will be displayed, showing two clusters. Users can further explore these results by exporting them into the native sheet experience.

The clustering capability in Qlik Sense enables users to uncover hidden insights and identify patterns within their data. By visualizing the relationships between different variables, users can make data-driven decisions and gain a deeper understanding of their business processes.

Feature Release: Key Drivers

In addition to the existing analysis types, we are excited to unveil a soon-to-be-released feature called key drivers. Key drivers allow users to identify and measure the impact of different influencers on a target variable. This feature is primarily used for modeling purposes and empowers users to understand what factors contribute to a specific outcome.

For example, let's consider the Scenario of wanting to influence customer churn. By using key drivers, users can determine the influencers that affect customer churn and analyze their impact. The influencers can be ranked using various scores, such as F1 score, to identify the most significant contributors. Armed with this knowledge, users can modify the data model and make data-driven changes to improve the target variable's outcome.

While key drivers are still under development, users can activate this functionality directly within the sheet experience. This allows frontline analysts to guide discussions and explore different scenarios, leading to a better understanding of the underlying factors and potential strategies for improvement.

Future Developments: Proactive Insights

Looking ahead, Qlik Sense is continually evolving, and one exciting development on the horizon is proactive insights. Proactive insights aim to deliver calculated insights asynchronously and in a consumable format. Users will receive valuable insights within their existing workflows, whether in sheets or through centralized feeds.

These insights will enable users to stay informed and make data-driven decisions effortlessly. Whether it is uncovering hidden patterns, identifying anomalies, or predicting future events, proactive insights will empower users to extract maximum value from their data. As this feature is still undergoing active development, stay tuned for more updates and details on how proactive insights will revolutionize analytics in Qlik Sense.

Conclusion

In this article, we explored the integration of machine learning services in Qlik Sense's analysis types. Leveraging machine learning, Qlik Sense enhances its predictive and decision-making capabilities, allowing users to uncover valuable insights and make data-driven decisions. Whether it is forecasting, time series decomposition, clustering, or anomaly detection, machine learning services empower users to extract Meaningful information from their data.

We also showcased the clustering capability within Qlik Sense, highlighting its ability to reveal hidden patterns and relationships within data. Additionally, we introduced the upcoming feature release of key drivers, which allows users to identify influential factors and measure their impact on a target variable.

Looking toward the future, Qlik Sense's proactive insights feature holds immense potential for delivering Timely and actionable insights to users. With this feature, users will be able to effortlessly access valuable information within their existing workflows, opening doors to new possibilities and greater data-driven decision-making.

For more information and updates on these features and others, be sure to explore our "What's New" section and reach out to our product management team.

Resources

• Qlik Help Center
• Qlik What's New
• Contact Qlik Product Management