Center for Machine Learning and Intelligent Systems
About  Citation Policy  Donate a Data Set  Contact


Repository Web            Google
View ALL Data Sets

Audiology (Standardized) Data Set

Below are papers that cite this data set, with context shown. Papers were automatically harvested and associated with this data set, in collaboration with Rexa.info.

Return to Audiology (Standardized) data set page.


Vassilis Athitsos and Stan Sclaroff. Boosting Nearest Neighbor Classifiers for Multiclass Recognition. Boston University Computer Science Tech. Report No, 2004-006. 2004.

We did not use four datasets (dermatology,soybean, thyroid, audiology because they have missing attributes, which our current formulation cannot handle. One dataset (ecoli) contains a nominal attribute, whichour current


Marcus Hutter and Marco Zaffalon. Distribution of Mutual Information from Complete and Incomplete Data. CoRR, csLG/0403025. 2004.

the traditional filter F, as well as the naive Bayes classifier, are now computed using the empirical probabilities (20). The remaining implementation details are as in the case of complete data. Data set #feat. FF F BF Audiology 69 64.3 68.0 68.7 Crx 15 9.7 12.6 13.8 Horse-colic 18 11.8 16.1 17.4 Hypothyroidloss 23 4.3 8.3 13.2 Soybean-large 35 34.2 35.0 35.0 Table 4: Average number of attributes


Richard Nock and Marc Sebban and David Bernard. A SIMPLE LOCALLY ADAPTIVE NEAREST NEIGHBOR RULE WITH APPLICATION TO POLLUTION FORECASTING. International Journal of Pattern Recognition and Artificial Intelligence Vol. 2003.

The average accuracy of sNN is bold-faced because a paired t-test reveals a threshold risk of order < 1/10% when comparing it w.r.t. NN and tNN (see text). Accuracy# Dataset NN tNN sNN Audiology 65.45 4.63 65.96 4.48 68.79 2.48 Australian 77.89 0.88 77.88 0.83 77.68 1.27 Balance 85.35 1.11 86.48 1.53 86.59 1.48 Bigpole 64.73 1.23 64.75 1.34 65.43 1.19 Breast-W 95.89


Alexander K. Seewald. How to Make Stacking Better and Faster While Also Taking Care of an Unknown Weakness. ICML. 2002.

with number of classes and examples, discrete and continuous attributes, baseline accuracy (%) and entropy in bits per example (Kononenko & Bratko, 1991). Dataset cl Inst disc cont bL E audiology 24 226 69 0 25.22 3.51 autos 7 205 10 16 32.68 2.29 balance-scale 3 625 0 4 45.76 1.32 breast-cancer 2 286 10 0 70.28 0.88 breast-w 2 699 0 9 65.52 0.93 colic 2 368


Wai Lam and Kin Keung and Charles X. Ling. PR 1527. Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong. 2001.

and their codes Data set Code Automobile Ab Auto-Mpg Am Audiology Au Balance-scale Ba Breast-cancer-w Bc Car Ca Credit screening Cs Ecoli Ec Glass1 Gl Hepati He Ionosphere Io Iris Ir Letter Le Liver Li Monk-1 M1 Monk-2 M2


Alexander K. Seewald and Johann Petrak and Gerhard Widmer. Hybrid Decision Tree Learners with Alternative Leaf Classifiers: An Empirical Study. FLAIRS Conference. 2001.

sizes of the trees generated on the complete training set. Table 2. Classification errors (%) and standard deviations for base learners: J48 with reduced error pruning, Linear, IB1 and NaiveBayes. Dataset J48-R Linear IB1 NaiveBayes audiology 25.44#1.87 20.93#0.98 21.90#0.56 27.79#0.65 autos 29.61#2.40 34.59#1.77 25.95#1.00 42.24#1.26 balance-scale 21.22#1.25 13.38#0.58 13.25#0.55 9.50#0.29


Jihoon Yang and Rajesh Parekh and Vasant Honavar. DistAl: An inter-pattern distance-based constructive learning algorithm. Intell. Data Anal, 3. 1999.

TABLE II Comparison of generalization accuracy between various algorithms. DistAl is the results of our approach and NN is the best results in [40]. Dataset DistAl NN Annealing 96.6 96.1 Audiology 66.0 77.5 Bridge 63.0 60.6 Cancer 97.8 95.6 Credit 87.7 81.5 Flag 65.8 58.8 Glass 70.5 72.4 Heart 86.7 83.1 Heart (Cleveland) 85.3 80.2 Heart (Hungary) 85.9


Mark A. Hall. Department of Computer Science Hamilton, NewZealand Correlation-based Feature Selection for Machine Learning. Doctor of Philosophy at The University of Waikato. 1999.

algorithm using the features in the subset, and then 111 evaluating its performance on the test set 4 . Figure 6.23 shows plots of CFS-UC's merit versus naive Bayes' accuracy on a selection of datasets (chess end-game, horse colic, audiology and soybean). Plots for the remaining datasets---and for when IB1 and C4.5 are used to measure accuracy---can be found in appendix E. The first thing that


Pedro Domingos. Unifying Instance-Based and Rule-Based Induction. Machine Learning, 24. 1996.

minimum of 4 examples (instead of 2) in two branches of a test, and use a confidence level of 37.5% for rule pruning (instead of 25%). The fine-tuned algorithms were then tested on the remaining 15 datasets in Table 4 (from audiology to zoology). Note that this procedure is somewhat unfavorable to RISE, since some of these domains were previously used in the development of the other algorithms, as


Geoffrey I. Webb. OPUS: An Efficient Admissible Algorithm for Unordered Search. J. Artif. Intell. Res. (JAIR, 3. 1995.

B. C. 465,058 * 1,211,211 * * * Execution terminated after exceeding the 24 CPU hour limit. # Only three of ten runs completed. Table 4: Number of nodes explored under best-first fixed-order search Data set Runs Minimum Mean sd Audiology 0 --- --- --- House Votes 84 10 451,038 1,319,911 624,957 Lenses 10 51 64 9 Lymphography 10 597,842 2,251,652 1,454,583 Monk 1 10 463 788 225 Monk 2 10 4,283 5,895 931


Thomas G. Dietterich and Ghulum Bakiri. Solving Multiclass Learning Problems via Error-Correcting Output Codes. CoRR, csAI/9501101. 1995.

as ``soybean-large'', the ``audiologyS'' data set as ` audiology standardized'', and the ``letter'' data set as ``letter-recognition''. 267 Dietterich & Bakiri 2.2 Learning Algorithms We employed two general classes of learning methods: algorithms


Jerome H. Friedman and Ron Kohavi and Youngkeol Yun. To appear in AAAI-96 Lazy Decision Trees. Statistics Department and Stanford Linear Accelerator Center Stanford University.

that have large differences. The LazyDT's average error rate is 1.9% lower, which is a relative improvement in error of 10.6% over C4.5's 17.9% average error rate. Three datasets deserve special discussion: anneal, audiology and the monk2 problem. Anneal is interesting because ID3 manages so well. An investigation of the problem shows that the main difference stems from


Alexander K. Seewald. Meta-Learning for Stacked Classification. Austrian Research Institute for Artificial Intelligence.

with number of classes and examples, discrete and continuous attributes, baseline accuracy (%) and entropy in bits per example (Kononenko & Bratko, 1991). Dataset cl Inst disc cont bL E audiology 24 226 69 0 25.22 3.51 autos 7 205 10 16 32.68 2.29 balance-scale 3 625 0 4 45.76 1.32 breast-cancer 2 286 10 0 70.28 0.88 breast-w 2 699 0 9 65.52 0.93 colic 2 368


Bernhard Pfahringer and Ian H. Witten and Philip Chan. Improving Bagging Performance by Increasing Decision Tree Diversity. Austrian Research Institute for AI.

procedure---either five five-fold cross-validations, or the use of a prespecified test set. D R A F T October 14, 1997, 7:30pm D R A F T IMPROVING BAGGING 11 Table 1. Characteristics of the used datasets. Dataset #Ex #C #Cat #Num Test Audiology 226 24 69 0 5 Theta 5 cv Breast 699 2 0 9 5 Theta 5 cv Colic 368 2 14 8 5 Theta 5 cv Credit 690 2 9 6 5 Theta 5 cv Diabetes 768 2 0 8 5 Theta 5 cv DNA


D. Randall Wilson and Roel Martinez. Improved Center Point Selection for Probabilistic Neural Networks. Proceedings of the International Conference on Artificial Neural Networks and Genetic Algorithms.

reduction in size can be even more dramatic when there are more instances available. This is especially true when the number of instances is large compared to the complexity of the decision surface. Dataset Anneal Audiology Australian Breast Cancer (WI) Bridges Crx Echocardiogram Flag Heart (Hungarian) Heart (More) Heart Heart (Swiss) Hepatitis Horse-Colic Iris Liver-Bupa Pima-Indians-Diabetes


Alexander K. Seewald. Dissertation Towards Understanding Stacking Studies of a General Ensemble Learning Scheme ausgefuhrt zum Zwecke der Erlangung des akademischen Grades eines Doktors der technischen Naturwissenschaften.

with number of classes and examples, discrete and continuous attributes, baseline accuracy (%) and entropy in bits per example (Kononenko & Bratko, 1991). Dataset cl Inst disc cont bL E audiology 24 226 69 0 25.22 3.51 autos 7 205 10 16 32.68 2.29 balance-scale 3 625 0 4 45.76 1.32 breast-cancer 2 286 10 0 70.28 0.88 breast-w 2 699 0 9 65.52 0.93 colic 2 368


Geoffrey I Webb. Learning Decision Lists by Prepending Inferred Rules. School of Computing and Mathematics Deakin University.

were compiled by M. Zwitter and M. Soklic at University Medical Centre, Institute of Oncology, Ljubljana, Yugoslavia. The Audiology data set was compiled by Professor Jergen at Baylor College of Medicine. References Clark, P., & Boswell, R. (1991). Rule induction with CN2: some recent improvements. In Proceedings of the Fifth European


Mohammed Waleed Kadous and Claude Sammut. The University of New South Wales School of Computer Science and Engineering Temporal Classification: Extending the Classification Paradigm to Multivariate Time Series.

Saito [Sai94], and further worked on by Manganaris [Man97]. The task is to classify a stream as one of three classes, cylinder (c), bell (b) or funnel (f ). Samples are generated as follows: 2 These datasets are: arrythmia, audiology bach chorales, echocardiogram, isolet, mobile robots, waveform. 6. Experimental Evaluation 161 c(t) = (6 + #) # [a,b] (t) + #(t) b(t) = (6 + #) # [a,b] (t) (t -


Mohammed Waleed Kadous. Expanding the Scope of Concept Learning Using Metafeatures. School of Computer Science and Engineering, University of New South Wales.

a custom learner works, but is labour-intensive. Relational learning techniques tend to be very sensitive to noise and to the particular clausal representation selected. They are typically 1 These datasets are: arrythmia, audiology bach chorales, echocardiogram, isolet, mobile robots, waveform. unable to process large data sets in a reasonable time frame, and/or require the user to set limits on the


Return to Audiology (Standardized) data set page.

Supported By:

 In Collaboration With:

About  ||  Citation Policy  ||  Donation Policy  ||  Contact  ||  CML