history markscheme
1- A friend of yours wants to create an expert system that will attempt to guess the results of football games based on the players that make up every team on their statistics. Explain clearly the stages, processes, and persons involved in the development of this expert system as well as how the system would work.
In the development of an expert system that would predict football matches of the
Game based on the football players that make up each team one would have to do a statistical research.
Football match results are affected by a number of reasons. this includes the winning and loosing probability, whether away or at home. Despite the accuracy of the results, the frequency of accuracy should be considered when making the expert system. This is due to the probability factor in a football match, which is caused by many factors of a football match. In the making of a prediction system the following have to be made. Consideration of a good feature set is an important stage in initial development of the system. A feature set might contain factors that help better predict the football match. This might include Players ratings; previous wins or losses and probably the history of achievement of both teams.
The feature set should consist of at least 20sets of match data for each feature. This improves the chances of prediction the more training data a system gets the more reliable it becomes. Some of the data that might be used for the feature sets required by the software engineer should be collected from FIFA (Federation International Football Association) as it reserves the biggest database of diverse features some of which an engineer might require.
The data has to be trained and tested on its accuracy this is called a testing and validation
Process. The first process is to allow a machine learning algorithm to process the data of the
Chosen feature set and calculate the probabilities. In this process we encounter a problem that some data
In the development of an expert system that would predict football matches of the
Game based on the football players that make up each team one would have to do a statistical research.
Football match results are affected by a number of reasons. this includes the winning and loosing probability, whether away or at home. Despite the accuracy of the results, the frequency of accuracy should be considered when making the expert system. This is due to the probability factor in a football match, which is caused by many factors of a football match. In the making of a prediction system the following have to be made. Consideration of a good feature set is an important stage in initial development of the system. A feature set might contain factors that help better predict the football match. This might include Players ratings; previous wins or losses and probably the history of achievement of both teams.
The feature set should consist of at least 20sets of match data for each feature. This improves the chances of prediction the more training data a system gets the more reliable it becomes. Some of the data that might be used for the feature sets required by the software engineer should be collected from FIFA (Federation International Football Association) as it reserves the biggest database of diverse features some of which an engineer might require.
The data has to be trained and tested on its accuracy this is called a testing and validation
Process. The first process is to allow a machine learning algorithm to process the data of the
Chosen feature set and calculate the probabilities. In this process we encounter a problem that some data
Bibliography: Extra Credits: The Uncanny Valley - YouTube. 2013. Extra Credits: The Uncanny Valley - YouTube. [ONLINE] Available at:http://www.youtube.com/watch?v=9K1Kd9mZL8g. [Accessed 23 September 2013]. Image File:Mori Uncanny Valley.svg - Wikipedia, the free encyclopedia. 2013.File:Mori Uncanny Valley.svg - Wikipedia, the free encyclopedia. [ONLINE] Available at: http://en.wikipedia.org/wiki/File:Mori_Uncanny_Valley.svg. [Accessed 23 September 2013].