Lecture# 5 (Part 1) Crystal Ball, Clairvoyant, Fortune Telling ... Can Predictive Analytic Deliver the Future ?

Question 1:

Many predictive analytic models are based on neural network technologies. What is the role of neural networks in predictive analytics? How can neural networks help predict the likelihood of future events. In answering these questions, specifically reference Blue Cross Blue Shield of Tennessee.

The role of the neural networks can act as replacement for analysts in retail, manufacturing and many other industries use a variety of statistical methods to solve a range of problems in forecasting, data classification and pattern recognition and produce forecasts as accurate as or better than those available from other statistical methods. In fact, neural networks offer many advantages, including: improved accuracy over traditional statistical methods; a unified approach to a wide variety of predictive analytics problems; and they requires fewer statistical assumptions and can manage complex predictive analytics tasks in a more automated way, which saves time for analysts and programmers

The likelihood of future events in neural networks are assisted by the fact that they can learn to adjust to new circumstances, lend themselves to parallel processing, function without having all the information or having that information in a structured format, copy with huge amounts to data with different variables, and analyze relationships found in the data Blue Cross Blue Shield of Tennessee has a neural network which they use to predict health resources that will be needed after certain procedures. The patterns they find can help determine if a patient will have a reaction to the procedure. Having this data quickly can help save health care costs and patients.

Question 2:

What if the Richmond police began to add demographic data to its predictive analytics system to further attempt to determine the type of person (by demographic) who would commit a crime. Is predicting the type of person who would commit a crime by demographic data (ethnicity, gender, income level, and so on) good or bad?

It is not a issue of good or bad to add demographic data to its predictive analytics system to further attempt to determine the type of person (by demographic) who would commit a crime by demographic data. The main issue here is whether there is any infringement of privacy law or whether the person who use the predictive system will use it in a proper manner but not to abuse it. 

Question 3:

In the movie Gattaca, predictive analytics were used to determine the most successful career for a person. Based on DNA information, the system determined whether or not an individual was able to advance through an education track to become something like an engineer or if a person should complete only a lower level of education and become a janitor. The government then acted on the system’s recommendations and placed people in various career tracks. Is this good or bad use of technology? How is this different from the variety of personal test you can take that informs you of your aptitude for different career?

 In my opinion, this is a bad use of technology by acted according to the system recommendation.  This is because everyone has different potential which yet to be displayed. We can not simply determine a future of a person by relying on a predictive system. 

The personal test is served as a reference for someone to know about the best suit career for them.  Basically it is no much different from the predictive system because it is also based on the data collected and recommend a best option to a person.

Question 4: What role can geographic information system (GISs) play in the use of predictive analytics? As you answer this question, specifically reference FedEx’s use of predictive analytics to:

(1)    Determine which customers will respond negatively to a price increase and

(2)    Project additional revenues from proposed drop-box locations.

A geographic information system (GIS) integrates hardware, software, and data for capturing, managing, analyzing, and displaying all forms of geographically referenced information. GIS allows us to view, understand, question, interpret, and visualize data in many ways that reveal relationships, patterns, and trends in the form of maps, globes, reports, and charts.  A GIS will helps us answer questions and solve problems by looking at our data in a way that is quickly understood and easily shared. GIS technology can be integrated into any enterprise information system framework.

 By using a GIS, FedEx can know whether their sales increase or drop in a particular area. When a price increase in  a particular area, the sale will usually drop and FedEx can trace the complains from customer regarding the price from a particular region. 

When a particular area has a high sales volume, FedEx can estimate more sale base on GIS. FedEx may increase the production from a particular area to generate additional revenue. 

Question 5:

The department of Defense (DoD) and the Pacific Northwest National Laboratory are combining predictive analytics with visualization technologies to predict the probability that a terrorist attack will occur. For example, suspected terrorists caught on security cameras who loiter too long in a given place might signal their intent to carry out terrorist attack. How can this type of predictive analytics be used in airport? At what other buildings and structures might this be used?

Predictive analytics is a way of using “Big Data” techniques to predict possible outcomes, from results of presidential elections to the behavior of individual people. Depending on the application, predictive analytic data often includes demographic information, family and marital status, purchasing history, past weather patterns, business transaction histories, social media activity, website clicks, and, of course, telephone metadata, all of which help shape the map of an possible underlying reality.

The same concept can be adopted here. The data of the criminal can be share in the airport and the data of a criminal such as which area of that particular criminal likelihood to appear and the travel data of the criminal to be shared in the system in the airport. When the CCTV detected the person, the security in the airport may be alarmed to be more alert.

This method can be used in anywhere and any building so long as the system properly installed and operated.  

Lecture # 5 (Part 2) - Decision Support is Good for Your Health

Lecture # 5 -
 Decision Support is Good for Your Health

Question 1:

The system discussed in this case was a decision support system. However, other types of computer-aided support are utilized in medicine. Can you think of ways that the medical profession could use AI system? For example, how about pattern recognition? Could that help in diagnosing illness?

By utilizing Multiple-agent systems, it can be used to predict the possible treatment available for the patients, estimate the effect of different medicine (such as chemotherapy) that may effective for certain type of diseases (such as cancer), anticipate the possible diagnosis, etc.

Pattern recognition is definitely a useful tool for medical industry. This is because nowadays one patient has many health problems. And it is impossible for human to collect all these data from each patient, to recognize the health complications patterns with naked eyes. Only by utilizing AI system such as data mining agents, it can utilize classification to finds patterns in information and categorizes them into those classes. Using the system, we can enter an abundant amount of patients’ medical history, thus making the statistical analysis more reliable and accurate.  And base on the highest occurrence of symptoms, doctor can know which disease is that.

Question 2:

A big worry in the collating and aggregation of medical information across department and even medical institution is that the more access there is to a person’s medication information, the more exposed that personal information becomes. HIPAA (Health Insurance Portability and Accountability Act), signed into law in 1996, addresses the security and privacy of your health data. The law was enacted to try to ensure that medical records, electronically stored and transferred, would be protected. Do you think that making your medical records available to the various branches of the medical industry (doctors, therapists, insurance companies, hospital biling, etc) is, on the whole, good or bad? Why? Can you think of any instances where disclosure of medical information could cause problems for a patient?

We think that it is necessary to make our medical records available to doctors, pharmacist, radiotherapist, etc because by understanding the full history of the patient, then only can the doctor or therapist make the most accurate diagnosis. There are many situations whereby the doctors haven’t thorough check the patient’s medical history, and give the wrong diagnosis to the patient. A typical example happen in cancer patient, whereby the doctor instruct chemotherapy without truly finding out the origin of the cancer, thus inducing wrong type of chemotherapy that cause permanent damage to the patient without curing the cancer disease. Therefore, making the medical records available to various relevant parties has its benefits.

Of course, it comes with the security issue for patients’ medical history. Therefore, it is important to take precaution steps, before the implementation of it. As long as you have a reliable system to safeguard the patients’ medical information, there should have no problem to implement this.

Possible problems for patient is that they can never cheat the insurance company about their medical condition, because everything will be transparently disclose to insurance agent.

Question 3:

Could predictive analytic be a part of the HHC decision support system? If so, what sort of data would it analyse? What might it tell medical staff? Would it be useful only to those who are already ill or could it help healthy people? How?

Yes of course predictive analytic can be part of the HHC decision support system. It collects all historical information to predict future events and outcomes. For example, the system will analyse what age group of people prone to what kind of chronic diseases, obese people prone to what kind of diseases, what type of disease prone to recurrence, what combination of medication are the best and most effective to cure certain disease, what are the possible causes for certain disease, etc.

It will definitely help not only the ill people, but also those health people. For example, healthy people also go through medical check-up. Sometimes when they have certain minor symptoms (example, having trace amount of blood in the urine which cannot be seen by naked eyes), doctors may not be able to identify the health problem immediately. By looking at the system with historical data, doctors may be able to find out the clues, thus speed up the treatment process, and helping the patient cured at the earlier stage. 

Question 4:

A clinical study has shown that telemonitoring, discussed briefly in this case, helps in keeping down medical costs. In fact, monitored patients were hospitalized about half as often as those with the same illnesses who were not monitored. Emergency room visits were five times more likely among those who were unmonitored. What types of illnesses could be monitored this way (think chronic diseases like high blood pressure)? Would it make sense to use the system as follow-up care? How could the data be utilized to help those who might become sick in the future? 

Telemonitoring involves easy-to-use equipment that helps you track your vital signs at home. Diabetes mellitus, hypertension and high cholesterol are vital to use the system as follow-up care. By knowing the signs and symptoms, the patients can be taught on the precaution steps to prevent severe condition happen. When patients’ blood glucose, cholesterol and blood pressure is extraordinary high, doctors can alert them, and teach them to practice healthy eating and lifestyle, creating the awareness for prevention. 

Question 5:

Could an automated medical diagnosis system ever replace live doctors? Why or why not? Would you trust an experienced doctor over a database that you could query yourself? Why or Why not? 

Automated medical diagnosis system can never replace live doctors because there could be unpredictable medical condition occurred, based on how the prior diagnosis is done. Therefore, a machine can never be able to respond to emergency situation like live doctors.

We cannot totally trust on experienced doctor because they are human, and human could make mistakes. As compared to database, the information generated from database is much more accurate than experienced doctor. However, we can never fully rely on database because it could have errors as well, as sometimes it may not be able to generate answer when the information has never exist before. Thus I think both database and experienced doctor is very important. We cannot lose either one of them. And with database, it can help to minimize the error that might be made by experienced doctor. Therefore, having a combination of these two will definitely induce synergistic benefits to the medical industry.