Data Mining Audiology Records with the Chi-Squared Test and Self-Organising Maps

In our project AudioMine we wish to address the problem that we need to understand more of the underlying factors influencing which patients would benefit from being fitted with a hearing aid. We describe some results from our pilot study, in which two data mining techniques, the chi-squared test and self-organising maps, were used to discover associations between various fields in 20,000 patient audiology records. We discuss methods of determining the degree of benefit experienced by hearing aid users, so in our main study, working with a larger data set, we will be search for associations between features of audiology records and degree of hearing aid benefit.

@InProceedings{OCW05, 
 	 author =  {Oakes, Michael Philip and Cox, Shaun and Wermter, Stefan},  
 	 title = {Data Mining Audiology Records with the Chi-Squared Test and Self-Organising Maps}, 
 	 booktitle = {22nd British National Conference on Databases},
 	 editors = {Jackson, Mike; Nelson, David; Stirk, Sue },
 	 number = {},
 	 volume = {},
 	 pages = {123--130},
 	 year = {2005},
 	 month = {},
 	 publisher = {University of Sunderland Press},
 	 doi = {}, 
 }