Fluent Essays

In summary follow the steps in the chapter to install WEKA and R software.WEKA https://waikato.github.io/weka-wiki/documentation/R http://cran.mirrors.hoobly.com/Enclosed are the chapter and tutorials:WEKA: https://www.slideshare.net/wekacontent/an-introduc...And https://www.cs.waikato.ac.nz/ml/weka/R tutorials: https://online.stat.psu.edu/statprogram/tutorials/...And http://www.r-tutor.com/r-introductionMake sure you create a document with your results for both software for credit. chapter3_weka_and_r.pdf introductiontoweka.pdf Unformatted Attachment Preview Beginning with Weka and R Language 3 Chapter Objectives To learn to install Weka and the R language To demonstrate the use of Weka software To experiment with Weka on the Iris dataset To introduce basics of R language To experiment with R on the Iris dataset 3.1 About Weka In this book, all data mining algorithms are explained with Weka and R language. The learner can perform and apply these algorithms easily using these well-know data mining tool and language. Let’s first discuss the Weka tool. Weka is an open-source software under the GNU General Public License System. It was developed by the Machine Learning Group, University of Waikato, New Zealand. Although named after a flightless New Zealand bird, ‘WEKA’ stands for Waikato Environment for Knowledge Analysis. The system is written using the object oriented language Java. Weka is data mining software and it is a set of machine learning algorithms that can be applied to a dataset directly, or called from your own Java code. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. The story of the development of Weka is very interesting. It was initially developed by students of University of Waikato, New Zealand, as part of their course work on data mining. They had implemented all major machine learning algorithms as part of lab work for this course. In 1993, the University of Waikato began development of the original version of Weka, which became a mix of Tcl/Tk, C, and Makefiles. In 1997, the decision was made to redevelop Weka from scratch in Java, including implementations of modeling algorithms. In 2006, Pentaho Corporation acquired an exclusive license to use Weka for business intelligence. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 29 This chapter will cover the installation of Weka, datasets available and will guide the learner about how to start experimentation using Weka. Later on we will discuss another data mining tool, R. Let us first discuss the installation process for Weka, step-by-step. 3.2 Installing Weka Weka is freely available and its latest version can be easily downloaded from https://www.cs.waikato. ac.nz/ml/weka/downloading.html as shown in Figure 3.1. Figure 3.1 Downloading Weka To work more smoothly, you must first download and install Java VM before downloading Weka. 3.3 Understanding Fisher’s Iris Flower Dataset R. A. Fisher’s Iris Flower dataset is one of the most well-known datasets in data mining research. The Iris dataset is commonly used in texts on data mining to illustrate various approaches and tools. Becoming familiar with this dataset will aid you in using the data mining literature to advance your knowledge of the subject. Fisher’s Iris dataset is available inside the ‘data’ folder of the Weka directory as ‘iris.arff ’ or you can find it at link http://archive.ics.uci.edu/ml/datasets/Iris and can be downloaded as shown in Figure 3.2. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 30 Data Mining and Data Warehousing Figure 3.2 Downloading the Iris dataset The origins of Fisher’s dataset can be traced to Anderson’s Iris dataset because Edgar Anderson collected the data to quantify the morphologic variation of Iris flowers of three related species. This dataset contains 50 samples of each of the three species, for a total of 150 samples. A sample of the Iris dataset is shown in Figure 3.3. Figure 3.3 Sample of the Iris flower [see colour plate] Anderson performed measurements on the three Iris species (i.e., Setosa, Versicolor, and Virginica) using four iris dimensions, namely, Sepal length, Sepal width, Petal length, and Petal width. He had observed that species of the flower could be identified on the basis of these four parameters. So, he prepared a dataset for its analysis. In data mining terminology, these four iris dimensions are termed as ‘attributes’ or ‘input attributes’. The three iris species are known as ‘classes’, or ‘output attributes’ and each example of an iris is termed as ‘sample’, or ‘instance’. A section of the Fisher’s dataset spreadsheet is given in Figure 3.4, showing the four input attributes and one output attribute, or class. This figure only shows five of the 150 instances, or Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 31 samples, of irises. For this dataset, the input attributes are numerical attributes, meaning that the attributes are given as real numbers, in this case in centimeters. The output attribute is a nominal attribute, in other words, a name for a particular species of Iris. Figure 3.4 Sample of Fisher’s dataset 3.4 Preparing the Dataset The preferred Weka dataset file format is an Attribute Relation File Format (ARFF) format. Weka also accepts several alternative dataset file formats, one of them being the Comma Separated Values (CSV) file format. Commonly dataset is available in XLS (Excel file) format, and in order to process excel data with Weka, the first step we need to do is convert our XLS file into a CSV file. For this open the dataset spreadsheet by using excel in XLS format, and then choose Save As, and then Other Formats. On this screen, we choose the file type as CSV as shown in Figure 3.5. When the information box appears, select Yes. In this way, we can open any XLS file in Weka by first converting it into CSV. Figure 3.5 Save as ‘Other Format’ Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 32 Data Mining and Data Warehousing 3.5 Understanding ARFF (Attribute Relation File Format) An ARFF file is an ASCII text file that describes a list of instances sharing a set of attributes. ARFF files were developed for use with the Weka machine learning software. ARFF files have two distinct sections. The first section is the Header information, which is followed by the Data information as shown in Figure 3.6. Figure 3.6 3.5.1 ARFF format of IRIS dataset ARFF header section The header of the ARFF file contains the name of the relation, a list of the attributes (the columns in the data), and their types. The first line in the ARFF file defines the relation and its format is given as follows. @relation Where is a string and if the relation name consists of spaces then it must be quoted. The declaration of attributes includes an ordered sequence of @attribute statements. Each attribute of the dataset is defined by @attribute statement in the ARFF file which uniquely defines the name and data type of attribute. The format for the @attribute statement is given as follows. @attribute Where the must start with an alphabetic character and if attribute-name consists of spaces then it must be quoted. The can be of any types such as numeric, string, date and nominal. The keywords numeric, string and date are case insensitive. Numeric attributes can Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 33 be real or integer numbers and string attributes consist of textual values. Date attribute is declared according to the format given as follows. @attribute date [] Where represents attribute name and is an optional string which represents that how date values should be printed. The default format string is ‘yyyy-mm-dd hh:mm:ss’ as shown below. @RELATION Timestamps @ATTRIBUTE timestamp DATE ‘yyyy-mm-dd hh:mm:ss’ @DATA ‘2018-05-15 10:15:10’ ‘2018-05-14 09:50:55’ Nominal values are defined by providing an listing the possible values: {, , , ...} as shown below. For example, the class value of the Iris dataset can be defined as follows: @ATTRIBUTE class {Iris-setosa, Iris-versicolor, Iris-virginica} The values that contain spaces must be quoted. The order sequence of declaration of the attributes indicates the column position of the attribute in the data section of the ARFF file. For example, if an attribute is declared at second position then Weka expects that all that attributes values will be found in the second comma delimited column. 3.5.2 ARFF data section The ARFF Data section of the file contains the data declaration line and the actual instance lines. The data declaration line consists of @data statement that is a single line representing the start of the data segment in the ARFF file. Each instance of the dataset is represented on a single line and the end of the instance is specified with carriage return. All the values of the attributes for each instance are delimited by commas and the missing values are represented by a single question mark. The format of the data section of ARFF file is given as follows. @data 1.4, 2.3, 1.8, 1.5, ABC The values of string and nominal attributes are case sensitive, and any that contain space must be quoted. For example, the ARFF format of IRIS dataset looks like as shown in Figure 3.6. 3.6 Working with a Dataset in Weka When you install Weka, you can expect to find a Weka icon on your desktop. When you start up Weka, you will see the Weka GUI Chooser screen as shown in Figure 3.7. From Weka Chooser, we can select any of the applications such as Explorer, Experimenter, KnowledgeFlow, Workbench and Simple CLI. The brief description about these applications is given in Table 3.1. In this chapter, we will cover the application Explorer only as we can directly apply data mining algorithms through this option. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 34 Data Mining and Data Warehousing Figure 3.7 Table 3.1 Weka GUI Chooser screen WEKA GUI applications Application Description Explorer It is an environment for exploring data. Experimenter This interface is for designing experiments with your selection of algorithms and datasets, running experiments and analyzing the results. Knowledge Flow It is a Java-Beans based interface to design configurations for streamed data processing. Workbench It is a unified graphical user interface that combines the other three such as Explorer, Experimenter and Knowledge Flow (and any plugins that the user has installed) into one application. Simple CLI It provides a simple command-line interface and allows direct execution of Weka commands. Figure 3.8 Weka Explorer screen Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 35 Select the Explorer application. Selecting the Explorer application displays the Weka Explorer screen as shown in Figure 3.8. Select the Preprocess tab to load the dataset to Weka. On the Preprocess tab, select Open file… and then select the FishersIrisDataset.csv file. Note that all six attribute columns in our dataset have been recognized, and that all 150 instances have been read as shown in Figure 3.9. For the purpose of analysis, remove the instance number attribute by selecting it in the Attributes check box, or by clicking on it, as it does not play any role as shown in Figure 3.9. Figure 3.9 3.6.1 Loading Fisher’s dataset Removing input/output attributes Since we have just removed an attribute, this would be a good point to save. Here, we choose to save our modified dataset in the preferred Weka.arff file format. As expected, we are now working on just 5 total attributes, having removed the ‘Instance’ attribute as shown in Figure 3.10. The Explorer Preprocess screen provides several types of information about our dataset. There are three main elements of this screen as shown in Figure 3.11. i. The class designator, ii. The attribute histogram, iii. The attribute statistics. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 36 Data Mining and Data Warehousing Figure 3.10 Fisher’s dataset after removal of instance number Figure 3.11 Elements of the Explorer screen Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 37 Expand the Class Designator box as shown in Figure 3.12. By default, the last (rightmost) column in the dataset is used as the output attributes, or class attribute. Here, class is by default considered as the output attribute to be used for classifying iris samples. Weka allows you to change the class attribute to any attribute in the dataset. Figure 3.12 3.6.2 Expansion of class designator Histogram To see a histogram for any attribute, select it in the Attributes section of the Preprocess tab. Figure 3.13 represents the histogram that shows us the distribution of Petal widths for all three species. As it turns out for this histogram, dark blue is Setosa, red is Versicolor, and bright blue is Virginica. The histogram shows that there are, for example, 49 samples in the lower histogram bin for Petal width, all of which are Iris-Setosa, and shows that there are 23 samples in the highest bin, all of which are Virginica. The histogram also shows that there are 41 samples in the middle bin, in which most of the samples belong to Versicolor irises and rest are Virginica. Now, click on the Visualize All (as shown in Figure 3.13) button to see the histograms of all the attributes together. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 38 Data Mining and Data Warehousing Figure 3.13 Figure 3.14 Histogram for Petal width [see colour plate] Histograms for all attributes of Iris dataset [see colour plate] Figure 3.14 shows histograms for all input attributes of the iris dataset. By comparing the histograms for all of the input attributes, we can begin to get a sense of how the four input attributes vary with different iris species. For example, it appears that Iris Setosa tends to have relatively small Sepal length, Petal length, and Petal width, but relatively large Sepal width. These are the sorts of patterns that data mining algorithms use to perform classification and other functions. Notice also that the species histogram verifies that we have 50 of each iris species in our dataset. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 39 3.6.3 Attribute statistics In statistics for attributes, for example in case of Petal width attribute as shown in Figure 3.15, we see that we have no missing data, in other words, there are no instances in the dataset which are without Petal width measurement. It also shows basic statistics of the selected attribute in the form of its minimum, maximum, mean and standard deviation values as shown in Figure 3.15. It also provides values for two characteristics termed Distinct and Unique. Figure 3.15 Attribute statistics [see colour plate] First, consider the Distinct characteristic. The distinct characteristic shows how many different values are taken on by a particular attribute in the dataset. In case of Petal width attribute, we see a segment of the Iris Dataset showing seven of the 150 samples in the dataset as shown in Figure 3.16. For just this segment of seven samples in the iris dataset, we see four distinct values for Petal width, i.e., 0.1, 0.2, 0.6 and 0.3. There are a total of 22 distinct values for Petal width in the entire dataset of 150 samples. The Unique characteristic on the Explorer screen tells us the total of Petal measurement values that appear only once in the full dataset, i.e. out of 150 samples. In the case of attribute ‘Petal Width’ we have three samples with Petal width of 0.1, 0.6 and 0.3 that are unique in selected instances of the dataset as shown in Figure 3.16. However, overall we have only 2 unique samples in the entire dataset of 150 samples as indicated in Figure 3.15. To practice this concept, let us find distinct and unique values for the following dataset: 23, 45, 56, 23, 78, 90, 56, 34, 90 Solution: Distinct: 6 (23, 45, 56, 78, 90, 34) Unique: 3 (45, 78, 34) Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 40 Data Mining and Data Warehousing Figure 3.16 3.6.4 Distinct and Unique values ARFF Viewer With the ARFF Viewer, we can view the attributes and data of a dataset without loading of the dataset. When the ARFF Viewer opens up, select Open, then find and open the Fisher’s Iris dataset file. Figure 3.17 (a) shows how to open the ARFF Viewer from Weka GUI Chooser and further, Figure 3.17 (b) shows how to open a file in ARFF Viewer. Figure 3.17 (a) Selecting ARFF Viewer from GUI Chooser and (b) opening the file in ARFF Viewer Note that Weka has identified the Sepal and Petal values as numeric, and species as a nominal attribute. The highlighting of the label ‘species’ shows that this is specified as the class attribute for the dataset. It is important to note that we can perform a number of operations on the dataset using the AARF Viewer’s File, Edit, and View functions. Figure 3.18 shows the ARFF Viewer of Fisher’s dataset. Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 01 May 2019 at 23:38:07, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/9781108635592.004 Beginning with Weka and R Language 41 Figure 3.18 3.6.5 ARFF Viewer of Fisher’s dataset Visualizer It is also possible to do data visualization on our dataset from the Weka GUI Chooser. On the GUI Chooser, choose Visualization, and then Plot as shown in Figure 3.19. Figure 3.19 Visualization of dataset Figure 3.20 shows the plot between Sepal length and Sepal width. Here, Sepal length is displayed on the x-axis, and Sepal width on the y-axis. Using the drop-down boxes, we can change the ... Purchase answer to see full attachment