Ch 12 BANAAAAAAAAAAAAAAAA :D Flashcards
What is bana
process of developing deciisons or reccomendations based on insights from the data
what is business intelligence
apps, technolgoeis, and prcoesses for gathering, storig, accessing and analyzing data to elp business users make better decisions
what do info systems do
support the managers job
3 roles of the manager job ACCORIDNG TO MINTZBERG
- interpersonal roles: figurehead, leader
- informational role: monitor, disseminator, spokesperson
- decisional roles: entrepreneur, disturbance handler
3 steps in deciison making
intelligence (identify problem, define opportunity)
design (construct a model for addressing situation)
choice (select a solution oro course of action to solve problem)
why do managers need IT support
too many alts
time pressure
need a sophisticated analysis
often necessary to rapidly access remote info, consult experts, or conduct a group decision making session
problem structure
Problem Structure: dtructured, semi-strucutre, unstructure
NATURE OF DECISIONS: 3 categories of managerial decisions (operational, management, strategic)
The first dimension is problem structure, in which decision-making processes fall along a continuum ranging from highly structured to highly unstructured (see the left column in Figure 12.2). Structured decisions deal with routine and repetitive problems for which standard solutions exist, such as inventory control. In a structured decision, the first three phases of the decision process—intelligence, design, and choice—are laid out in a particular sequence, and the procedures for obtaining the best (or at least a good enough) solution are known. These types of decisions are candidates for decision automation.
At the other extreme of complexity are unstructured decisions. These decisions are intended to deal with “fuzzy,” complex problems for which there are no cut-and-dried solutions. An unstructured decision is one in which there is no standardized procedure for carrying out any of the three phases. In making such a decision, human intuition and judgement often play an important role. Typical unstructured decisions include planning new service offerings, hiring an executive, and choosing a set of research and development (R&D) projects for the coming year. Although BA cannot make unstructured decisions, it can provide information that assists decision makers.
Located between structured and unstructured decisions are semistructured decisions, in which only some of the decision-process phases are structured. Semistructured decisions require a combination of standard solution procedures and individual judgement. Examples of semistructured decisions are evaluating employees, setting marketing budgets for consumer products, performing capital acquisition analysis, and trading bonds
nature of deicisons
The second dimension of decision support deals with the nature of decisions. All managerial decisions fall into one of three broad categories:
Operational control: Executing specific tasks efficiently and effectively
Management control: Acquiring and using resources efficiently in accomplishing organizational goals
Strategic planning: The long-range goals and policies for growth and resource allocation
the decision matrix
The three primary classes of problem structure and the three broad categories of the nature of decisions can be combined in a decision-support matrix that consists of nine cells, as diagrammed in Figure 12.2. Lower-level managers usually perform the tasks in cells 1, 2, and 4. The tasks in cells 3, 5, and 7 are usually the responsibility of middle managers and professional staff. Finally, the tasks in cells 6, 8, and 9 are generally carried out by senior executives.
Today, it is difficult to state that certain organizational information systems support certain cells in the decision matrix. The fact is that the increasing sophistication of ISs means that essentially any information system can be useful to any decision maker, regardless of their level or function in the organization. As you study this chapter, you will see that business analytics is applicable across all cells of the decision matrix.
3 targets of bana
- devloping one/many related analytics applications
- developing infrastructure for enterprise wide analytics
- support for org transformation
what are pain points
business problems
what do we need to do bana
underlying technologies!! we need a lot of data with high transmission speed
what is data management
we need to clean data and get into data marts and data warehouses through ETL
bana tools
excel
OLAP (multi deminsional analysis)
data mining
dss
regression
what is descriptive analytics
first step in data reduction- it summarizes what has happened in the past, enables decision makers to learn from past behaviours
BA tools for descriptive analytics
OLAP
Data mining
dss
-sensitivty anlaysis
-what if analysis
-goal seeking analaysis
olap: slicing and dicing data (either drill down or roll up) USE DATA CUBES
data mining: identifying patterns and predicting trends in large database
DSS:mathematical models to analyze semistructred & unstructured problems
-sensitivity: how sensitive to change
-what if: concerning uncertainty
-goal seeking: backward solution approach
descriptive analytics apps names
fandango
optumRx
Predictive analytics meaning
examine recent and historical data to detect patterns and predict future trends
pPredictive analytics BA Tools
data mining
linear regression
Data mining: use vast amounts of data to predict the future!! ex:targeted marketing, forecasting levels of bad loans, tracking crime patterns
linear regresison: using data to predict futurre with statistics
example of predictive analytics paplications
FANDANGO: sending emails to people who might be interested in concerts similar to what they went too
websites predicting which ads youll click on
leading online dating sites!
consequences of preditcitvie analytics UNINTENDED
-Target: preicted woman was pregnant and started iving ad
=Uber: RAISING rates too high as trying to get away from a shooter
how do descriptive analytics and predictive analytics work together to bridge gap between data management and actionable busienss deciison
DATA MANAGEMENT:
To investigate this business problem, you collect data on service calls for the past year from the CAASCO locations within your district. You first “clean” the data by adjusting for outliers and missing values, resulting in a final sample of 3,219 service calls.
You are now ready to build a regression model using temperature as the independent variable to predict the number of customer service calls received on any given day; this is the dependent variable. You will use the results to predict how many employees should be available to receive and dispatch customer service calls.
DESCRIPTIVE ANALYTICS: or this example, the correlation between daily low temperature and the number of service calls was found to be –0.84, with an average of 48 service calls per day. The correlation is negative because as the temperature decreases, the data indicate that the number of service calls increases.
The square of this correlation, R2 = .71, is the predictive power of the model. That is, the model, using low daily temperature, explains approximately 71 percent of the variation in the number of calls received each day. Therefore, 1 minus R2 means that 29 percent of the variance in the dependent variable is due to extraneous or unexplained variables.
PREDICTIVE ANALYITSC: The manager has decided to use linear regression for the predictive analysis. To do so, certain reasonable assumptions must be met:
There must be at least 30 data points.
The relationship between the independent and dependent variables must be linear. The linearity assumption can best be tested with scatterplots.
Even though the data are assumed to be normally distributed, the manager should check this assumption.
The sample size of 3,219 satisfies the first assumption. The manager then used Excel to test the second assumption, producing a scatterplot to determine if the plot of each ordered pair of data (independent variable, dependent variable) produced a linear pattern. The scatterplot for the data did exhibit a linear pattern. Therefore, linear regression is an acceptable statistical procedure for these data.
To check for normality, the manager used the Ryan-Joiner normality test to calculate the correlation between the data and the normal scores of the data. If the value is near 1, then the sample dataset is likely to be normal. This dataset met the normality assumption.
Now that the three linear regression assumptions have been met, the next step is to define the linear regression model between these two variables using Excel or a similar statistical package. The linear regression model is:
Number of calls received = 124.79 – 1.5 × (daily low temperature)
These results indicate that for one degree of increase in the daily low temperature, the predicted daily number of calls received will decrease by 1.5. That is, CAASCO will expect to receive fewer calls on warmer days and more calls on colder days. At a temperature of 0 degrees (x = 0), the expected number of calls will be approximately 125.
ACTIONABLE BUSINESS:Based on the linear regression, the district manager is able to use the projected daily low temperature for up to 14 days in advance to predict how many service calls the offices will receive each day. (The Weather Network provides reasonably accurate daily temperatures on a 14-day outlook.) Therefore, the manager can predict how many employees will be needed to manage the expected number of service calls in order to ensure low wait times for the customers.
ASK NEXT QUESTION: t this point, the manager can return to the data management stage with new input variables. For the CAASCO data in this example, it is feasible to consider another business problem with the appropriate inputs or to expand the analysis by considering other variables relevant to the business question in the example. For example, the manager might want to include the actual time of day, by hour, so that staffing levels could be more accurately decided. The manager also might want to examine the location of the CAASCO branches as a variable. Adding these variables would require the use of new regression models.
You also want to recall that our example is for the South Central Ontario district, where temperatures are warmer than in some other regions of the country. Therefore, if we were to predict data in the northern regions, then we would have to build a different model based on the data from the North.
Our example proceeds from data management, to descriptive analytics, to predictive analytics (through a simple linear regression model). We address how the results of predictive analytics often lead to additional questions that include additional variables, which would require a multiple linear regression model.
From a statistical perspective, we might ask: Aren’t there many different analytical approaches to solving the same problem? The answer is yes. But a more important question to ask is: Which one approach is the best? The answer to this question is—none! The best approach depends on the kind of data you are working with. And, because data come in all shapes and sizes, there cannot be one best approach for all problems. Therefore, selecting the best model for the particular data is always an important exercise in data analytics.
Be
What is prescriptive abalytics
recommends a course of action, identified likely outcome of each deciison
BA tools used in prescirptive analyics
fandango: using the most desirable movie time to maximize profit by raising price
google driverless car: ccar anyicipating pedestrians to not hit them
presentation tools
-dashboards
-GIS
dashboards: simple and easy access to timely information, create reports
gIS: is a computer-based system for capturing, integrating, manipulating, and displaying data using digitized maps. Its most distinguishing characteristic is that every record or digital object has an identified geographical location. This process, called geocoding, enables users to generate information for planning, problem solving, and decision making. The graphical format also makes it easy for managers to visualize the data. There are countless applications of GISs to improve decision making in both the public and private sectors.
For example, when the Toronto Police Service decided to develop a tool to improve the understanding of policing, increase transparency, and enhance confidence in the Service, it found that using geospatial data can help it accomplish this mission better. The result was a public safety data portal build using GIS software from ESRI (www.esri.ca).
The tool provides a centralized platform to easily access useful and timely crime data. A wide range of crime information, including shootings, homicides, and major crime indicators, are included in the app. In addition, users can select a set of dates, types, and neighbourhoods based on which the app will query the dataset and visually depict the results on a map. This app is used by various users, such as researchers, media outlets, university professors, students, and residents concerned about public safety in their neighbourhood. The Vancouver Police Department has also developed a similar app using GIS software from ESRI.
These apps help visualize crime hot spots and enable crime analysts to study the reasons for certain areas being targets of criminal activities. Therefore, the apps enable the police to develop crime prevention strategies and tactics in order to reduce the number of crime incidents. In addition, these apps can increase police transparency and accountability as they shed light on the key activities and crime issues that the police are engaged with.
bana and bi are interchangeable
word
what are dashboards
provide easy access to timely info and direcct access to management reports
(user friendly, craphics, management can drill down orscale up)
strategic/tactical dashboards
-SHOW KPIS, and analyze historical data and visualize trends
operational and informational dashboards
understand events, projects ro asset statuts
inform and engagee audiences
what is a network operations centre
a centralized location where computer, telecommunications or satellite networks systems are monitored and managed 24-7
CONTROL ROOM
WHAT DOES THE CPS THING HERE SHOW
VISUALIZATIONS
