Week 10 Question: Due March 29
Answer one of the following questions as a comment on this post.
Submit your contribution by 11:00 AM on Thursday, March 29. It should be insightful (don’t just repeat back facts from the class), but it doesn’t need to be more than a few sentences. Most of all, I want to know what you think!
- Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
- Take a look at the SpotCrime website. Type in a few zip codes to see crime statistics in different areas. Would you consider this more an example of data analytics as we’ve discussed it this week, or more like the OLAP analysis we have been doing over the last week or so? Explain your reasoning.
David Schuff
#2
The statistics that SpotCrime retrieves is both OLAP and data analytics we’be discussed this week. The crime statistics reveals constantly updated crimes that happen in specific times and locations (OLAP analysis). This same data can be used to detect geographic patterns (crime more prevalent in cities), peak times for crimes (more frequent during evening) and can use this information to send police officers to be located at the most frequent crime areas in different times of the day to lower crime and efficiently expend police resources.
Question #1: I remember one of my good examples is when I have to keep track of all tutor’s database. My job duty was to maintain all the records, so I have to figure out what kind of academic subjects did the tutors help the most. This is because to find which kind of specific tutors we should hire, I need to do some analyzing first. Therefore, I believe this kind of example represents a descriptive analytic, because without a good description I cannot depended on the estimation at the same time.
1.Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
After doing some outside research music data mining could be very useful for a number of reasons. Finding patterns and similarities in songs not only allows us to place them into specific genres, but we can also use data mining to find particular patterns between those invdividuals who listen to certain music. Also, because it has been said before that certain people listen to certain types of songs or artists before they committ crimes. Music data mining is an example of predictive analytics because we would be gathering information to make predictions about people and look for patterns to recommend music to people, like iTunes does.
1) Insurance companies always predict the probability of a claim. For example, the premium for auto insurance is based on the historic customer data. The company would want to use data mining and create a prediction model to find out the probability of accidents that specific customers will get into based on their history of accidents. If a customer’s past reveals that he/she got into 3 accidents in one year, then the premium that will be given to them will most likely be higher than someone who got into less accidents in the same period of time. This is an example of predictive analytics because the insurance company predicts the frequency of accidents that one would get into based on past performance.
#2. The SpotCrime website shows us the number of theft, shooting, burglary, etc, that have happened in a specific zipcode. This type of information is part of OLAP, but which at the same time, can be part data analytics. The reason is because all the information needed to do the analysis is there; it just needs to be put together. For instance, if Philadelphia police wants to put more police officers around the most dangerous areas, they already have the data of where most of the crime occurs. Thus, they can put them where they are needed.
#2
This is more like the OLAP analysis we did before, and it belongs to decision support. Dimensional database provides a retrospective, dynamic data delivery at multiple levels; it reports what crimes (shooting, theft, or arson) are happening now or what have happened before (time), and where crimes events are located. Anyway, it does not tell why these events are happening, and it does not predict what will happen as well.
1.) Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
An example of a problem where dating mining is useful is when colleges are sorting through applications for acceptance. Universities use data mining to analyze specific data about a student to decide whether that student would be a good fit for their program. For instance, they typically will analyze the students transcript and analyze which classes a student has completed and the grades they received. Also, their cumulative GPA is another determining factor as well as their participation in extra curricular activities. I think this would be an example of predictive analytics because it is classifying whether a student is granted or denied their acceptance to the college based on their data.
#1
Data mining can be useful if a company is trying to launch a new product and needs to determine its target customers. By carefully selecting the dimensions of the analysis for example: age, demographics, income, marital status, a marketing company can define the future target consumers and success for the new product. This is a predictive analytics example. Which is the likelihood that the product will be accepted in the market? And which customers are more likely to buy it?
After visiting the SpotCrime website the information that they provide could be considered OLAP. The site provides you with information that relates to decision support. What is happening is the type of crimes that are occurring such as robberies, thefts, assaults, etc. The time and date of the crime is also provided which allows visitors to know when these events have occurred. Data analytics with this site would involve the law enforcement agencies using this information to help reduce crime in certain areas.
# 2
I believe that spotcrime.com is a mixture of both data analysis as well as OLAP. It is data analysis in the sense that by looking at different zip codes and its crimes, one is coming up with conclusions about the area based on the statics given. It represents OLAP because it gives a detailed map about what has happened. The purpose of spotcrime.com is to represent the things that happened in the past not why they happened or predictions of what will happen in the future.
#2.
I think this is more of an example of OLAP, although it could be argued that it is also data analytics. It is an example more of raw data that while it is useful in its raw form, has not been analyzed to any extent. It simply returns events that have occurred on record. It would be different if it actually analyzed the crimes and you could see what times robberies occur or when you could expect to see a future crime happen. The site doesn’t give that extra analytical layer.
# 2
This website http://www.spotcrime.com is an example of OLAP. OLAP can tell you what is happening, or what has happened. The website tells you what is happening in the particular area and what has happened on particular date. Its not data analytic because its not really showing how much robberies or other typed of crimes occurred during a particular time period rather its only listing the different crimes on particular dates. We can’t really analyze the data as its showing on the website.
#1
Data mining can be used for super stores such as Wal-Mart and target to determine what types of inventory they should order. The amounts and types of items ordered can depend on store location, customers, seasons, and other demographics. This information can be very useful for these stores so they are not overstocked or understocked in certain inventory. This data can be used when determining what items consumers typically buy before a major snow storm or any other sever weather. This can help a store to better prepare by ordering a large amount of items that they know customers will tend to purchase during that time. Additionally this information will help store to prepare for upcoming seasons and events when ordering specific inventory.
#2
I think that the SpotCrime website is a combination of the OLAP and data analytics. It is partly an OLAP because of the constant updates to the database. The results for every hour or even every minute can be different, making this website similar to the transactional database with instant access. On the other hand, the outcomes from the database are summarized by the zip codes. This process can be based just on few filters that extract only certain parts of data from the whole database. However it can be a really simple task, it summarizes and analyses the raw facts, presenting information, not a raw data. Thus, this website serves as both, utilizing the advantages of OLAP and data analytics.
The SpotCrime website shows OLAP analysis because it shows what has happened at a specific location, the time of the crime, and what type of crime it was (assault, robbery, shooting). However, this type of information can be used in data analytics. Data analytics takes the raw data and turns it into useful information. For example, police officers can understand crime patters and managers can schedule them accordingly. Maybe there are 0 crimes before 10pm and 5 crimes between 11pm and 2am. Therefore, instead of having a full staff during the day, they can schedule more officers at night and less in the day to be more productive and efficient.
Question #2
The SpotCrime website is an example of both OLAP analysis and data analytics. SpotCrime shows OLAP analysis by constantly updating criminal events which occur at specific locations and times. However, the information displayed on the website could also be used for data analysis. By taking this raw data (where certain crimes occur and when) authorities could discover trends and similarities about these crimes. This analysis could in turn allow police to strategically place officers in certain locations, at certain times, which could significantly reduce criminal activity.
1. Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
An example of data mining is what Apple is currently doing with it’s iTunes application. They have added a feature in the newer iTunes software called Genius, by selecting certain songs and activating the feature it will allow Apple and the iTunes software to gather information about your music preferences and make playlists and recommendations according your selections. This would be an example of both descriptive analytics and predictive, Apple is attempting to categorize its customers based on their preferences in music and generating a recommendation based on some algorithm that would predict what other songs similar their customer would also enjoy.
Question #2
After going on the SpotCrime website, I believe that the SpotCrime is OLAP system, because the website provides us the historical data of the crimes in the real time data and certain location. It gives us the raw information about crimes. This site does not predict where and when the crimes may happen in the future. Therefore, it is not an example of data analysis, but it is a OLAP analysis.
The SpotCrime website should be an OLAP analysis. Through the zip code, we are able to determine historical data bout crimes each the area. These statistics will able for us to explain why a certain event is happening, and help predict what will happen in the future.
An example of a problem where data mining would be useful would be when a company is trying to figure out the best location to set up shop. The company would need to use data such as age, gender, social class, demographics, etc., to best match the location with their target market.
Data mining is the process of analyzing data from different perspectives and summarizing it into useful information that can be used to increase revenue, cuts costs, or both. It can be used to find correlations or patterns among dozens of fields in large relational databases. For example, supermarket chain can used the data mining to analyze local buying patterns. If they discover that many customers will buy soda and juice on Saturdays, they need to make sure that there are enough in the shelves, or they may put some kinds of soda near the entrance to remind customers and to make soda easy to take for them. This would be an example for descriptive analytics.
Question 2.
I think spotcrime is an example of OLAP because it doesnt give information on why each crime happened. If it did, then it’d be an example of data analytics. It tells what happened not why the crimes happened. OLAP is only good for viewing present and past events and cannot predict the future, this is what spotcrime does.
The Spotcrime Web site is an example of OLAP analysis. Because when I type in the zipcode to search crimes, the result I got is a real time data. I can slice the data to find out what specific crimes took place in a specific area by putting in a zipcode, therefor, the web site is an OLAP analysis but not a data analysis.
Question 2.
Data mining is an essential tool in the credit card industry. It can help identify fraud when pieces of information overlap or repeat over the millions of applications and card holders. This a description method because it looks at the existing data to find patterns within the data that might resemble fraud. Such as small purchases less than a dollar followed by large purchases of alcohol or electronics. It is also a great tool for trend analysis in buying habits as well as payment habits. This would be used primarily as a prediction method to anticipate credit card default risks and expected payment habits to anticipate interest income.
#1
Data mining would be useful in the company called JustFabulous which is a woman’s shoes, bags, and accessories website. When signing up for the site they try to get to know their potential cutomer by asking multiple questions about different styles and colors the person likes so they can put together a personal boutique for that specific person. So every month based off of a customers anwsers to the initial questions at registration and the items they have purchased, JustFabulous gathers information about your style so that they can make better recommendations for your next months styles. This is an exammple of predictive analytics because they use the style survey to initially put together a customers style based off of what they select and descriptive analytics once the customer begins to purchase different products JustFabulous can predict what future products will be similar in style to the previous selections.
1.Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
Data mining could be useful in tracking spending habits in past years to determine the cash/credit usage ratio in America and then investigate if seasonal holidays or weather patterns. Using descriptive analytics of cash flow and credit-card data from past years in companies like Walmart and credit data from Visa or AMEX, a trend can be plotted representing the annual amounts spent in each way. Following this trend, predictive analytics would suggest future sales ratios using the formulas derived from past data.
2. SpotCrime is useful for looking up information regarding what type of crimes occur, where they occur, and when they occur. This can be considered to be an OLAP system. This is considered OLAP because the website describes different crimes and when they occur, which are raw data, but this cannot be directly analyzed. The system does not do any analyzing for you, it just provides information in a general form.
1. Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
Data mining offers numerous methods for its user. For example, if a company such as Wal-Mart wanted to target a certain demographic, it would extract data from its database and attempt to make it useful information. It will help companies execute plans to target this demographic and cater to the demographic’s wants or needs in a product.
Question #1
Data mining would be very useful in the gaming/gambling industry. As we have previously noted in class while talking about databases, gaming companies collect humongous amounts of data regarding their casinos’ patrons. They use methods like gaming membership cards to track their gamers’ actions. When money is spent by a particular indivual on, say, a slot machine, the company knows about it. While I believe this paricular type of data mining is both predictive and descriptive, I think it is more of the former (Predictive). Casinos use the data collected to predict the movements/actions of their customers. They can cater to the gamers’ behavior. Ultimately, this leads to increases in profit.
An example where data mining would be useful is if you wanted to know what kind of sales to expect when you introduced one of your products that was already being sold but that had not yet been in this new market. This would be an example of descriptive analytics because you could look at past data patterns and apply them appropriately in the upcoming market you are entering.
1.Give an example of a problem where data mining would be useful. Then explain whether it would be an example of predictive or descriptive analytics (refer to the slides for the definitions if you don’t remember the difference).
At the chiropractic office I work for, we use patients’ forms to determine how they became aware of our office. Was it through our website or another patient? This helps us to distinguish where to market our services. This is an example of descriptive analytics.
1) A good example of data mining is the entire concept of the online radio player Pandora. Pandora tries to find trends and key indicators in your music listening habits to determine what songs you would like and what songs you wont like. Things like genre, year the music came out, whether you gave a song a “thumbs up” or “thumbs down” are all factored into whether or not they think you will enjoy a certain song or style of music. This would be a predictive analysis because Pandora is trying to guess your enjoyment of the next song they are gonna play, based on your previous preferences.
#2 the statistics shown on the spotcrime website show an in-depth analysis of crimes in any region of the country. However, this data is gathered by incidents, similair to transactions in a grocery store, each purchase can resemble a crime, hence why this site uses more of an OLAP system. You can recognize trends easily from showing where higher concentrations of crimes are so, in a sense there are uses of data analytics as well.
Question 1.
An example of data mining could have been used in my internship from this past summer when i worked with the Hurricane Junior Golf Tour. We had a database full of names, emails and phone-numbers of each junior golfer in the state of georgia, florida, and South Carolina. Each was held in a separate document based on the state but if we were able to analyze it and put in how many tournaments each player played in and where they played the company would have been able to focus marketing strageies on people who played less instead of the regular players. THis would be both descriptive and predictive because the database tells me where and when they live and play in our tournaments and then we would use that information to predict what future tournaments certain golfers would be most interested in.
Q1
Data mining is often used in the pharmaceutical industry. It is important to determine whether patients will react positively or negatively to the medicine. In addition, it would also be valuable to understand which factors or attributes of the patient caused these reactions. By performing studies on a group of patients, they are able to acquire the data needed to find the patients who reacted negatively and analyze whether there is a correlation between the groups attributes and the results. The same can be done for patients who reacted positively to the medicine and that information can be used to create a prediction model which would be used to target new patients. This is an example of predictive analytics as the pharmaceutical industry is using acquired data to predict which patients would be best suited for their medicine.
SpotCrime.com would be an example of online analytical processing. The website tells you what is happening or what has happened. It doesn’t show predictions or explain why something is happening the way it is. With data analytics, the website would have to provide more information to users. Right now, it is allowing users to make their own analysis when they see heavier crime in one area versus another.
2) Spot Crime website is an example of an OLAP analysis. It updates criminal activities ie. arson, burglary, theft, etc. in real time marked by the location. It also offers historical data and allows you to view these activities from previous months. However, it doesn’t offer predictions on what can or will happen in the future or gives an explanation on why these events are happening which would be consist with data analysis.
#1 An example of a problem where data mining would be useful would be looking at sales for different geographical regions. This explain is descriptive analytics because its a summary and its measure sales.
2. Take a look at the SpotCrime website. Type in a few zip codes to see crime statistics in different areas. Would you consider this more an example of data analytics as we’ve discussed it this week, or more like the OLAP analysis we have been doing over the last week or so? Explain your reasoning.
I typed in 3 zip codes which were in areas in which I lived that I wasn’t aware of so many crimes being done happening nearby and around me. The zip codes I entered were 19138, 19144, and 19119. Which are mainly Germantown and Mt. Airy areas of Philadelphia. However they seem to have all high theft comparison when doing an analysis of all three of them.
19138 zip code:
• Theft – 20
• Burglary – 10
• Assault – 6
• Robbery – 9
• Other – 3
19144 zip code:
• Theft – 43
• Burglary – 3
• Robbery – 3
19119zip code:
• Theft – 25
• Burglary – 14
• Assault – 4
• Other – 1
When doing this analysis it is very similar to what we have been doing in class if we would have included this data to analyze it more from a pivot table to transfer that data into a data cube to be able to see the different dimensions of the crimespots during which periods of time, and how long.
1. Data mining is would be useful for a company deciding to open a new retail location. Descriptive analytics could be used to determine where the ideal customers (target market) are located. Predictive analytics could be used to determine how many more customers could be gained through opening in a new location. For example a clothing company could look at past customers home addresses and determine where they may have a lack of reach.