Collab Decision Making and Portfolio Opt

Digital Analytics

Pricing and Revenue Management

Programming for Analytics I

Supply Chain Analytics

Visualization for Analytics

8 Figure 8: Travel-time and Travel-distance Selection Dialog Box After the travel time is selected, the algorithm is run using Esri ArcGIS Online. A polygon is drawn on the map based on the distance that can be covered in a given time on ground. If you select multiple travel times (such as five minutes and ten minutes), multiple polygons are drawn, with each one representing a different travel time. See Figure 9. 9 Figure 9: Travel-time Polygons for Five-Minute and Ten-Minute Drive Times CUSTOM REGIONS ON GEO MAPS In SAS Visual Analytics, you can render custom regions on a geo map. You can use the user interface in SAS Visual Analytics to import and register custom shape files. See Figure 10. See SAS Global Forum paper SAS1732-2018, “ Leverage Custom Geographical Polygons in SAS® Visual Analytics ” , by Falko Schulz. Figure 10. Using Custom Polygons on a Geo Map GEO CODING This capability is planned for Visual Analytics in future releases post 8.2 version. When you need to place points on a geo map, you will need the Latitude and Logintude coordinates for the points. I refer to this process of obtaining these Lat & long coordinates for the data points as Geo coding. For example, when you have the address of different customers to be displayed on a geo map, you want to convert the addresses to Lat & Longs. There are different approaches to doing the geo coding. One such method is to perform geo coding at a table as part of the Data Preparation. When you do geo coding on the table, the Lat & Long information is added to the table as additional columns and saved with the table. Any one who is try to use this table for analysis, reporting or exploration can use the geo coded data. 10 CONCLUSION In today ’ s business world, data is multiplying, and location information is available through IOT devices, smartphones, wearables, and other smart sensors and devices. Because so many of these new location data sources are available, being able to use this data together with other business data will provide better and deeper insights into how and why a business performs in a region. Location analytics will be an important component of business analysis, and SAS Visual Analytics can help customers use the business data and location data seamlessly. SAS Visual Analytics provides a range of advanced analytics such as forecasting, clustering, correlation, network, and text analytics. Using these analytics together with location-specific analytics will extend the use and reach of analytics in any organization. REFERENCES Schulz, Falko. 2018. “Leverage Custom Geographical Polygons in SAS Visual Analytics . ” Proceedings of the SAS Global Forum 2018 Conference. Cary, NC: SAS Institute Inc. Available http://support.sas.com/resources/papers/proceedings18 SAS Institute Inc. 2017. SAS® Visual Analytics: Working With Report Content . Cary, NC: SAS Institute Inc. Available: http://go.documentation.sas.com/?cdcId=vacdc&cdcVersion=8.2&docsetId=vaobj&docsetTarget=n14ezqx eg2a195n1au1q980gbkte.htm&locale=en SAS Institute Inc. 2017. SAS® Visual Analytics: Working with Report Data . Cary NC: SAS Institute Inc. Available: http://go.documentation.sas.com/?cdcId=vacdc&cdcVersion=8.2&docsetId=vareportdata&docsetTarget=p 031vp9uc5y5iun0zipy3c1trkqn.htm&locale=en#p1a7fw0vj2w6aln18sl77suu45o3 SAS Institute Inc. 2017. SAS® Visual Analytics 7.4: Administration Guide . Cary NC: SAS Institute Inc. Available: https://support.sas.com/documentation/cdl/en/vaag/69958/PDF/default/vaag.pdf SAS Institute Inc. 2017: SAS® 9.4 Programming, SAS/GRAPH: Mapping Reference . Cary NC: SAS Institute Inc. Available: http://go.documentation.sas.com/?cdcId=pgmmvacdc&cdcVersion=9.4&docsetId=grmapref&docsetTarget =p0qsj9oazola04n10usle21dvk67.htm&locale=en SAS Institute Inc. 2017: SAS Viya ™ 3.3 Administration: Data . Cary NC: SAS Institute Inc. Available: http://pubshelpcenter.unx.sas.com:8080/test/?docsetId=caldatamgmtcas&docsetTarget=p1dwawsidsczlp n121j0glleicxp.htm&docsetVersion=3.3&locale=en ACKNOWLEDGMENTS I would like to thank Jeff Phillips and Eric Short for their assistance about details about geo analysis and their critical role in improving the geographical visualization component and associated map algorithms.

DNSC | 6403 | Visualization for Analytics | study guides | 8 Figure 8: Travel-time and Tr

Visualization for Analytics [Week 1] Suneel Grover, M.B.A., M.S. The George Washington University (Dept. of Decision Science - M.S. in Business Analytics) - Professorial Lecturer SAS® Advisory Solutions Architect Before We Begin… https://www.youtube.com/watch?v=5_bDh5fV8Ok Agenda 1. Introduction 2. Syllabus Review 3. Lecture: Visualization for Analytics 4. Class Software Introduction ❖ SAS Visual Analytics ❖ SAS Visual Statistics ❖ SAS Visual Data Mining & Machine Learning (VIYA) How I Got Introduced To Visualization for Analytics There Is An Opportunity With This Class Experience https://youtu.be/POJ1ugMYPY4?list=PLRk0B9IRZe WiixBr2SQoQzROAGefuYsWM Course Syllabus – Let’s Walk Through It My Approach To Teaching ▪ Readings & Videos – We will have engaging conversations at the beginning of every class . ▪ Technical Software Assignments – There will be hands-on work to complement the readings and videos. ▪ Blackboard – I use it heavily for communications and homework assignments. Please make sure you keep an eye on it. ▪ Class Sessions - 30% theory / 70% technical skill development Reading Material Textbook Summary ▪ The role data visualization and infographics play in our world ▪ Basic principles of data and scientific reasoning ▪ How to become a better critical thinker ▪ Processes that will help you evaluate any data visualization (including your own) ▪ How to create and use effective charts, graphs, and data maps to explain data to any audience Visualization for Analytics Overview Data visualization is the presentation of data in a pictorial or graphical format It enables decision makers to see analytics presented visually, so they can grasp difficult concepts or identify new patterns With interactive visualization, you can take the concept a step further by using technology to drill down into charts and graphs for more detail, interactively changing what data you see and how it’s processed One Of My Favorite Quotes “The Greatest Value Of A Picture Is When It Forces Us To Notice What We Never Expected To See.” John W. Tukey, Exploratory Data Analysis 1977 Static Data Visualization Static visualizations are commonly seen as infographics posted on the web or printed as handouts. Usually focused on a specific data story, users can’t go beyond a single view to explore additional stories beyond what’s in front of them. The story is specifically captured in an engaging single page layout.

DNSC | 6403 | Visualization for Analytics | study guides | Visualization for Analytics [W

DNSC 6214: Pricing and Revenue Management Spring 2019 CASE 2: Cambridge Software Corporation Guidelines: This is a group work and is due March 25 th right before class starts. You only need to submit one report with the names of your group members; however you each need to be prepared to discuss it in class. Your report should answer the following two questions without exceeding five pages. You should also provide the necessary analysis supporting your arguments in an appendix (which will not be included in page count). One group member can upload your work on Blackboard. 1. If Cambridge Software Corporation offers only one version of Modeler, which version should it offer? 2. Should the firm offer more than one version of Modeler? If so, which versions should it offer? At what prices?

DNSC | 6214 | Pricing and Revenue Management | study guides | DNSC 6214: Pricing and Revenue

# PART A: ```{r} odev <- read.csv("SalesPerformance.csv") head(odev) ``` # The data is introduced to R ```{r} model1 <- lm(Proﬁt ~ Dist + Area + Popn + Outlets + Commis, data=odev) summary(model1) ``` # The R squared is 0.662 which means that the model explains 66.2% of the results. # The p-values of Area and Commis are smaller than 0.5 which means that they are signiﬁcant to the model. Outlets and Area have bigger p-values which means that making a type 1 error is high if this model is used to explain proﬁts. We should omit some variables to see which model gives the best result. ```{r} model2 <- lm(Proﬁt ~ Dist + Area + Popn + Commis, data=odev) summary(model2) ``` # Dropped outlets (highest p-value) but the R squared decreased which means the previous model was beer. So let’s try proﬁts/outlets as our dependent variable. ```{r} model3 <- lm(Proﬁt/Outlets ~ Dist + Area + Outlets + Popn + Commis, data=odev) summary(model3) ``` # The new model has more signiﬁcant variables so this model works beer. # PART B: ```{r} library(car) library(carData) vif(model3) cor(odev) ``` ```{r} model4 <- lm(Proﬁt/Outlets ~ Area + Outlets + Commis, data=odev) summary(model4) ``` # Area and Popn have high correlaon which can cause mullinearity. Also, Popn has the highest vif so I dropped them both. They also have higher p values. # All variables are important to the model and this gives the best ﬁt. # PART C: ```{r} plot(model4) ``` # The more linear the relaonship between dependent and independent variable, the beer the model. # Furthermore, the residuals of the regression should be normally distributed. This can be observed by looking at the Q-Q plot. ```{r} yeniodev <- odev[-c(19,32,47), ] yeniodev1<-lm(Proﬁt/Outlets ~ Area + Outlets + Commis, data=yeniodev) plot(yeniodev1) summary(yeniodev1) ``` # The above stated problems are parally solved with removing the 3 outlier rows from the data. # The R squared of the data went up which means that the model explains more. PART D: # The above calculaons show that rather than using only proﬁt as an indicator of success, we should use proﬁt per outlet since it explains more of the success with more variables. The ﬁnal model is Proﬁt/Outlets = 11.6677 – (0.2151*Area) – (0.02184*Outlets) + (2.1659*Commis). The model suggests that Area and Outlets are negavely correlated with the success. A thousand square mile increase in the area results in a 0.2151 loss of proﬁt per outlet. Also, if the outlet number increases by 1, the proﬁt per outlet decreases by 0.02184. Also, having a full-commission representave adds 2.1659 proﬁt per outlet compared to a parally commissioned representave. Furthermore, the model has a standard error of 0.7894 which means that 95% of the values predicted by this model fall in between +/- 2*0.7894=1.5788 proﬁt per outlet deviaon from the actual values. These 3 variables and the model itself have small p-values suggesng that the chances of this model explaining the success of a representave randomly is really low. Populaon and districts are removed from the model because they do not explain the performance of a sales representave well.

DNSC | 6302 | Programming for Analytics I | assessment | # PART A: ```{r} odev <- read.

Start ‐ up   document   for   the   SQL   exercise     This   handout   is   will   show   you   how   to   load   data   into   your   SQL   tables.   To   use   this   handout   you   must   also   have   access   to   the   myscript   text   file   that   contains   the   SQL   code.     As   you   can   see   from   the   first   screenshot   below,   I   have   only   one   file   ( myscript )   in   my   week06   folder.   Make   sure   that   the   myscript   file   in   the   folder   from   where   you   start   mysql.   The   command   to   start   mysql   is       mysql   –u   root  ‐ p     The   –u   option   is   followed   by   a   space   followed   by   the   name   of   the   user   (in   this   case   root).   The   –p   option   followed   by   nothing   tells   the   command   line   interpreter   that   the   user   is   going   to   provide   the   password   (see   below).                         Once you provide the password and press enter, your screen will look like what is shown below. Type out this command source myscript and press <enter>       The   contents   of   the   myscript   file   are   shown   below.   You   can   edit   that   file   and   check   the   statements   out   on   your   own.   SHOW DATABASES; DROP DATABASE IF EXISTS classwork; CREATE DATABASE classwork; USE classwork; SHOW TABLES; CREATE TABLE Student ( stuId CHAR(6), lastName CHAR(20) NOT NULL, firstName CHAR(20) NOT NULL, major CHAR(10), credits SMALLINT DEFAULT 0, CONSTRAINT Student_stuId_pk PRIMARY KEY (stuId), CONSTRAINT Student_credits_cc CHECK ((CREDITS>=0) AND (credits < 150)) ); INSERT INTO Student VALUES ("S1001", "Smith", "Tom", "History", 90); INSERT INTO Student VALUES ("S1002", "Chin", "Ann", "Math", 36); INSERT INTO Student VALUES ("S1005", "Lee", "Perry", "History", 3); INSERT INTO Student VALUES ("S1010", "Burns", "Edward", "Art", 63); INSERT INTO Student VALUES ("S1013", "McCarthy", "Owen", "Math", 0); Assume that all goes well, you should see this screen at the end of the execution of the scri p t. INSERT INTO Student VALUES ("S1015", "Jones", "Mary", "Math", 42); INSERT INTO Student VALUES ("S1020", "Rivera", "Jane", "CSC", 15); CREATE TABLE Faculty ( facId CHAR(6), name CHAR(20) NOT NULL, department CHAR(20) NOT NULL, rank CHAR(10), CONSTRAINT Faculty_facId_pk PRIMARY KEY (facId) ); INSERT INTO Faculty VALUES ("F101", "Adams", "Art", "Professor"); INSERT INTO Faculty VALUES ("F105", "Tanaka", "CSC", "Instructor"); INSERT INTO Faculty VALUES ("F110", "Byrne", "Math", "Assistant"); INSERT INTO Faculty VALUES ("F115", "Smith", "History", "Associate"); INSERT INTO Faculty VALUES ("F221", "Smith", "CSC", "Professor"); CREATE TABLE Class ( classNumber CHAR(8), facId CHAR(6) NOT NULL, schedule CHAR(8), room CHAR(6), CONSTRAINT Class_classNumber_pk PRIMARY KEY (classNumber), CONSTRAINT Class_facId_fk FOREIGN KEY (facId) REFERENCES Faculty (facId) ON DELETE NO ACTION ); INSERT INTO Class VALUES ("ART103A", "F101", "MWF9", "H221"); INSERT INTO Class VALUES ("CSC201A", "F105", "TuThF10", "M110"); INSERT INTO Class VALUES ("CSC203A", "F105", "MThF12", "M110"); INSERT INTO Class VALUES ("HST205A" ,"F115", "MWF11", "H221"); INSERT INTO Class VALUES ("MTH101B", "F110", "MTuTh9", "H225"); INSERT INTO Class VALUES ("MTH103C", "F110", "MWF11", "H225"); CREATE TABLE Enroll ( classNumber CHAR(8), stuId CHAR(6), grade CHAR(2), CONSTRAINT Enroll_classNumber_stuId_pk PRIMARY KEY (classNumber, stuId), CONSTRAINT Enroll_classNumber_fk FOREIGN KEY (classNumber) REFERENCES Class (classNumber) ON DELETE NO ACTION, CONSTRAINT Enroll_stuId_fk FOREIGN KEY (stuId) REFERENCES Student (stuId) ON DELETE CASCADE ); INSERT INTO Enroll VALUES ("ART103A", "S1001", "A"); INSERT INTO Enroll VALUES ("HST205A", "S1001", "C"); INSERT INTO Enroll VALUES ("ART103A", "S1002", "D"); INSERT INTO Enroll VALUES ("CSC201A", "S1002", "F"); INSERT INTO Enroll VALUES ("MTH103C", "S1002", "B"); INSERT INTO Enroll VALUES ("ART103A", "S1010", NULL); INSERT INTO Enroll VALUES ("MTH103C", "S1010", NULL); INSERT INTO Enroll VALUES ("CSC201A", "S1020", "B"); INSERT INTO Enroll VALUES ("MTH101B", "S1020", "A");

All courses in DNSC

Popular Documents