Other models
by Ferran Garcia Pagans, Neeraj Kharpate, Philip Hand, James Richardson, Dr. Henric
Qlik Sense: Advanced Data Visualization for Your Organization
- Qlik Sense: Advanced Data Visualization for Your Organization
- Table of Contents
- Qlik Sense: Advanced Data Visualization for Your Organization
- Credits
- Preface
- 1. Module 1
- 1. Qlik Sense® and Data Discovery
- 2. Overview of a Qlik Sense® Application's Life Cycle
- 3. Empowering Next Generation Data Discovery Consumers
- 4. Contributing to Data Discovery
- 5. Authoring Engaging Applications
- 6. Building Qlik Sense® Data Models
- 7. Qlik Sense® Apps in the Cloud
- 8. Extending the Qlik® Analytic Platform
- 9. Administering Qlik Sense®
- 10. Sales Discovery
- 11. Human Resource Discovery
- 12. Travel Expense Discovery
- 13. Demographic Data Discovery
- 2. Module 2
- 1. Getting Started with the Data
- Introduction
- Extracting data from databases and data files
- Extracting data from Web Files
- Activating the Legacy Mode in Qlik Sense® desktop
- Extracting data from custom databases
- Invoking help while in the data load editor or the expression editor
- Previewing data in the Data model viewer
- Creating a Master Library from the Data model viewer
- Using a Master Library in the Edit mode
- 2. Visualizations
- Introduction
- Creating Snapshots
- Creating and adding content to a story
- Adding embedded sheets to the story
- Highlighting the performance measure in a bar chart
- Associating persistent colors to field values
- Using the ColorMix1 function
- Composition
- Relationships
- Comparison
- Distribution
- Structuring visualizations
- 3. Scripting
- Introduction
- Structuring the script
- Efficiently debugging the script
- Packaging the code in script files
- How to use sub routines in Qlik Sense®
- Optimizing the UI calculation speed
- Optimizing the reload time of the application
- Using a For Each loop to load data from multiple files
- Using the Concat function to store multiple field values in a single cell
- 4. Managing Apps and User Interface
- Introduction
- Publishing a Qlik Sense® application created in Qlik Sense® desktop
- Creating private, approved, and community sheets
- Publishing a Qlik Sense® application to Qlik Sense® cloud
- Creating geo maps in Qlik Sense®
- Reference lines in Sales versus Target gauge chart
- Effectively using the KPI object in Qlik Sense®
- Creating Tree Maps
- Creating dimensionless bar charts in Qlik Sense®
- Adding Reference Lines to trendline charts
- Creating text and images
- Applying limitations to charts
- Adding thumbnails – a clear environment
- Navigating many data points in a scatter chart
- 5. Useful Functions
- Introduction
- Using an extended interval match to handle Slowly Changing Dimensions
- Using the Previous() function to identify the latest record for a dimensional value
- Using the NetworkDays() function to calculate the working days in a calendar month
- Using the Concat() function to display a string of field values as a dimension
- Using the Minstring() function to calculate the age of the oldest case in a queue
- Using the Rangesum() function to plot cumulative figures in trendline charts
- Using the Fractile() function to generate quartiles
- Using the FirstSortedValue() function to identify the median in a quartile range
- Using the Declare and Derive functions to generate Calendar fields
- Setting up a moving annual total figure
- Using the For Each loop to extract files from a folder
- Using the Peek() function to create a currency Exchange Rate Calendar
- Using the Peek() function to create a Trial Balance sheet
- 6. Set Analysis
- Introduction
- Cracking the syntax for Set Analysis
- Using flags in Set Analysis
- Using the = sign with variables in Set Analysis
- Point in time using Set Analysis
- Using comparison sets in Set Analysis
- Using embedded functions in Set Analysis
- Creating a multi-measure expression in Set Analysis
- Using search strings inside a set modifier
- Capturing a list of field values using a concat() function in Set Analysis
- Using the element functions P() and E() in Set Analysis
- 7. Extensions in Qlik Sense®
- 8. What's New in Version 2.1.1?
- Introduction
- Using the visual exploration capability in Qlik Sense® 2.1.1
- Defining variables in Qlik Sense®
- Exporting stories to MS PowerPoint
- Using the Qlik Dev Hub in Qlik Sense® 2.1.1
- Using Extension editor in Qlik Dev Hub
- Using Qlik Dev Hub to generate mashups
- Embedding Qlik Sense® application on a website using a single configurator
- Using the Qlik DataMarket
- Creating dynamic charts in Qlik Sense®
- Using Smart Search
- Using smart data load profiling
- Conclusion
- A. Appendix
- 1. Getting Started with the Data
- 3. Module 3
- 1. Getting Ready with Predictive Analytics
- 2. Preparing Your Data
- 3. Exploring and Understanding Your Data
- 4. Creating Your First Qlik Sense Application
- 5. Clustering and Other Unsupervised Learning Methods
- 6. Decision Trees and Other Supervised Learning Methods
- 7. Model Evaluation
- 8. Visualizations, Data Applications, Dashboards, and Data Storytelling
- 9. Developing a Complete Application
- A. Bibliography
- Index
In this section, we will see other models provided by Rattle in the Model tab, which aren't supervised learning. These methods are Linear and Logistic Regression, Neural Networks, and Survival Analysis.
Linear Regression is a statistical method to describe the relationship between one or more input variables and one or more output variables. The objective is to create a formula that models the relationships between input and output variables; in this way, we can use this formula to predict new observations.
Imagine you are the manager of a marina, your marina has a gas station and you need to predict the amount of gas oil you will sell during a summer day. On the Mediterranean coast, during the summer, the amount of gas oil sold is correlated to the temperature. The reason is obvious, on sunny days, the temperature rises and more tourists want to use their boats. The example is illustrated in this diagram:
Based on past experience, we know that on a day with a temperature of 26ºC, the gas station sold 3,000 liters, and on a 28ºC day, it sold 4,500 liters. In this example, we can assume that the relationship between Gas oil Sold and Temperature is described by this equation:
This equation describes the relationship between Gas oil Sold and Temperature. Using this equation and a weather forecast, we can predict the gas oil our gas station will sell tomorrow.
When our dataset has one output variable, we will call it Simple Linear Regression. If we have more than one output variable, we call it Multiple Linear Regression.
We use Linear Regression when the target variable is numerical. In classification tasks, when the target variable is categorical, we will use Logistic Regression. When the target variable has two possible values or classes, we can use binary or binomial Logistic Regression.
The Neural Network model is a method that can be used as supervised or unsupervised learning. Neural Networks are especially useful for pattern recognition and time series prediction; real-world applications include facial recognition, character recognition, or stock prices.
Neural Networks are inspired by the human brain and have three main layers:
- Input Layer: This layer receives the input data and passes them to the hidden layer.
- Hidden Layer: This layer contains a number of interconnected nodes. These nodes, called neurons, are mathematical functions that create the predictions.
- Output Layer: This layer creates the final prediction from the predictions done in the hidden layer.
Neural Networks work especially well, when the number of input or attributes in the observations is high. An important disadvantage of Neural Networks is that the Hidden Layer is a black box and the algorithm doesn't explain the value of the prediction.
-
No Comment