Neural Networks for Poker?!

http://www.pokerbot-smart.com/

Clicking on the link above will bring you to a website for Smart Poker Bot. This is an automated bot that is designed to play poker games online. The application collects information from many sources in order to increase your chances of winning. It takes into consideration the hand that you have, the cards on the table, your table position, and the actions of the other players. All the data is combined and stored in a neural network that predicts the hand strengths of your opponents.

Being able to know the strength of the players, you can play the game better. You can then decide how much to bid and how much you are willing to risk based on the provided information. The application makes real-time adjustments with each action taken during the game. When it is your turn to make a move, you have the most accurate information available to help you win (or cheat) in online poker.

You can leave the bot to play for you on its own. The advanced poker software can make human decisions based on the playing style of the user.

You can click on the software screen shots on the website to get a better understanding of how the application works. There are also videos on YouTube that show the application in process.

Microsoft Surface Demo for Patient Consultation

This video shows the introduction of a Microsoft service application made for the healthcare industry. It is an interactive, multi-touch screen application that helps doctors explain complex medical procedures to patients. It also is a way for patients to communicate important information to their doctors.

Both patients and doctors just need to place their healthcare card on the surface, and the application recognizes who they are and information about them. When doctors log on, they are given choices such as test results, reports, etc. When the patient logs on, they have choices such as x-rays, prescriptions, family history, etc. On the screen there is a display of the human body which can be altered to look at the different systems. Doctors and patients can work together to share and talk about x-rays. The doctor can magnify pictures, show videos, and use the human body to help explain procedures.

This is a great idea. The touch screen is fascinating and the ability of the cards to recognize the users and provide the correct information is great. I am only concerned with the privacy issue. Although the patient’s information is in a secure system called Health Vault, I am concerned about losing the card. Maybe they will advance it to fingerprints along with the card. Anyway, this is a great advancement and can help clarify certain procedures.

The Semantic Web

This is a great video that shows an introduction to the semantic web. It reminds me of the recommended video on the syllabus about the history of BI. Both videos are very easy to understand and have a light, comical feel to them.

The video clearly explains the difference between semantics and syntax. Syntax is how you say something and semantics is the meaning. The difference between these terms plays a big role in the current limitation of the web. The internet let users communicate with each other and exchange information. The World Wide Web led to quick and easy retrieval of documents and the html language. The problem was that the computers only understood the html language and just mimicked the info for retrieval. The computers could not realize the syntax (or the meaning) of the information being presented.

The semantic web helps computers understand the meanings of things on the web. They can also learn relationships between items. This helps the end user because if the computer can understand the meaning of what you are looking for, it won’t have to rely completely on key words to search for it. If it can understand what you looking for, the web and search engines can be more beneficial and successful in finding the best information. This can greatly save time for the end-user.

Reducing Information Overlaod with Text Mining

KnewCo is a company that has recognized the information overload on the web and has developed a platform with multiple applications that enables a leap in knowledge development. KnewCo recognized that the redundancy, inaccuracy, and blur of information on the web is the barrier to innovation and discovery.

The platform, The Concept Web, uses the process power of computers and the creative power of humans. It can distinguish between facts and statements and even recognize concepts. Once a concept is recognized, the platform assigns each a unique identifier and then builds a comprehensive view around each one. The platform also applies advanced algorithms to show predictive relationships between concepts. The ultimate goal of this company is to be able to provide a unique identifier for every concept in every language.

Once the relationships are formed, the platform can help with the information overload. The units of related concepts are called Knowlets. They show the changing the relationships between each concept. The advantage of the Knowlets is that when you search for a particular item, it will provide you with links for the most current knowledge on the topic. That way, you don’t have to sort through multiple sources when trying to find the most recent information on a topic.

KnewCo has many other applications that help with controlling the information you receive on the web. This platform is great in managing the knowledge of the exponential and uncontrollable expansion of the information on the web.

Data-Mining for Fingerprint Identification

This video explains how fingerprinting could use an upgrade to data mining technologies. A UC Davis law professor says that the reliability of fingerprint identification has declined while the population of the world continues to grow at a fast rate. The uses of digital cameras are a great upgrade, but they capture fewer details then the old traditional cameras. He suggests the use of data mining to go through the fingerprint collections to find patterns and correlations. This would enable workers to refine the objective criteria so that the computers could better identify the prints.

Tony Doonan, a state fingerprinting specialist, believes that there are only few mistakes in accurately identifying fingerprints. Failing to match prints is not because of the camera, but because of human error when taking the prints.

While it may be human error, I believe the use of data mining could help them make matches quicker, or at least be able to narrow done suspects more accurately. This is an interesting topic. I’m not entirely sure how the system works, but I do know that sorting and analyzing billions of fingerprints could benefit from the use of data mining.

Microsoft MapPoint: Geographic Cluster

This blog shows an example of a cluster analysis for data mining. Microsoft MapPoint is a program that can allow users to plot locations of objects and events. Although this program is not meant for cluster analysis, it can be used for that purpose. You just plug the location addresses into the program, and plotted screen can help you see possible relationships.

The example map provided shows John Snow’s famous map of the 1854 Broad Street Cholera Outbreak in London. By plotting the outbreaks, John was able to identify that the outbreak was centered around a specific water pump. The outbreak almost ended immediately when John realized this and removed the handle off that pump.

Businesses can now use this program for other business benefits. One example would be to analyze customer locations. If you find a pattern in the location of customers, you can use that cluster information to investigate why, focus marketing in that area, and identify future markets with the same characteristics.

The program lets you use different symbols and colors for plotting different items. Businesses could also use the cluster analysis to plot delivery times at different locations. This would enable businesses to better estimate shipping times for customers. This would result in better customer satisfaction.

http://www.mapping-tools.com/info/samples/5_cluster.shtml

Coral Reefs: GIS Example

A geographic information system (GIS) is a computer-based system that captures, stores, analyzes, and displays geographically referenced data by using digitized maps. I was a little confused on what kinds of companies would use this type of system, so I searched for an example. I found a coral reef company that used a GIS in order to determine a selection for MPA’s (Marine Protected Areas). The GIS was based on a variety of factors that enhance resilience to coral bleaching.

You can view a series of GIS maps that they used by clicking on the link http://www.reefresilience.org/pdf/coral_gis.pdf. Each screen shows an added factor that helped them determine the MPA. The factors include biodiversity value, physical factors, strong coral recruitment, and connectivity. All the data was combined and overlaid in order to enable managers to identify sites with the greatest potential to survive a bleaching event.