write what do you think about their writing. don’t use he/she
1. The following hyperlink depicts a case-study done by Microsoft to help deal with click fraud detection and prevention: http://www.appliedaisystems.com/papers/ClickQualitySystems54_LNCSFormat_clean.pdf (Links to an external site.)
The datamining techniques used by adCenter involve tracking,
association, clustering, and prediction. Through the various sources and
data collected, it was determined that detecting and dealing with click
fraud is extremely difficult as the various options have some
weaknesses. For example: checking events by IP address creates a
computational load much higher than is considered reasonable at 4
billion with IPv4 and 3×1038 with IPv6.
Having rules to analyze and interact with the IP and relevant data
coming through adCenter allows for accountability of all tasks being
performed due to the rule ID, but can lead to statistical problems in
efficacy if the rule hierarchy is poorly managed. A realistic example
would be: an interaction flags two rules: Blacklist and BotDetect, Depending on the hierarchy of the rules used to determine which fires first may result in it appearing that the BotDetect feature isn’t working well as the Blacklist rule takes precedence.
2. Data
mining is a useful field for finding usable information out of a large
database, and its usefulness has spread to many fields. This case study
talks about how data mining has become useful in the medical field.
Medical data is often very large and therefore difficult to understand
and extract information from. However, with data mining techniques, the
large data, with its many commonalities and outliers, can be better
understood. Some data mining techniques used include multilayer
perception, neural network, J48 decision tree, KNN, K-clustering, and
random forest.
For example, using neural networking has allowed medial researchers
to imitate and study the the actual behaviors of neurons. Decision Trees
are useful as well, as they allow for help to generate rules as well as
notice outliers. This helps with diagnosis and better understanding of
disorders and dieases.
Chirag, et all (2017, Nov-Dec). Data Mining Techniques in Healthcare:
a case study. International Journal of Computer Science Trends and
Technology(IJCST). Volume 5 issue 6.
http://www.ijcstjournal.org/volume-5/issue-6/IJCST…
3. there are 2 video attached. write what do you think about their presentation.
0 comments