Data
colleges_aaup

colleges_aaup

active ARFF Publicly available Visibility: public Uploaded 29-09-2014 by Joaquin Vanschoren
0 likes downloaded by 3 people , 3 total downloads 0 issues 0 downvotes
Issue #Downvotes for this reason By


Loading wiki
Help us complete this description Edit
Author: Source: Unknown - Date unknown Please cite: The AAUP dataset for the ASA Statistical Graphics Section's 1995 Data Analysis Exposition contains information on faculty salaries for 1161 American colleges and universities. The data may be obtained in either of two formats. AAUP.DATA contains the raw data in comma delimited fields with a single data line for each school. The order of variables is the same as given below for the fixed column version, although the spacing varies for each school. AAUP2.DATA has the data arranged in fixed columns, with two data lines for each school and a maximum line length of 80 characters. This dataset is taken from the March-April 1994 issue of Academe. Thanks to Maryse Eymonerie, Consultant to AAUP, for assistance in supplying the data. Faculty salary data are for the 1993-94 school year. You may wish to consult a copy of the special issue of Academe for more detailed descriptions of the variables. Data Revised: Wed Jan 18 1995 VARIABLE DESCRIPTIONS (AAUP2.DAT) Fixed column format with two data lines per school Line #1 1 - 5 FICE (Federal ID number) 7 - 37 College name 38 - 39 State (postal code) 40 - 43 Type (I, IIA, or IIB) 44 - 48 Average salary - full professors 49 - 52 Average salary - associate professors 53 - 56 Average salary - assistant professors 57 - 60 Average salary - all ranks 61 - 65 Average compensation - full professors 66 - 69 Average compensation - associate professors 70 - 73 Average compensation - assistant professors 74 - 78 Average compensation - all ranks Line #2 1 - 4 Number of full professors 5 - 8 Number of associate professors 9 - 12 Number of assistant professors 13 - 16 Number of instructors 17 - 21 Number of faculty - all ranks Missing values are denoted with * All salary and compensation figures are yearly in $100's To obtain the dataset from Statlib, send one of the single line messages below to the address statlib@lib.stat.cmu.edu send aaup.data from colleges or send aaup2.data from colleges For more information on the ASA Statistical Graphics Section's 1995 Data Analysis Exposition send the message send readme from colleges %%%%%%%%%%%%%% INFORMATION % %%%%%%%%%%%%%% WHAT'S WHAT AMONG AMERICAN COLLEGES AND UNIVERSITIES? This is the subject of the 1995 Data Analysis Exposition sponsored by the Statistical Graphics Section of the American Statistical Association. The purpose of the Exposition is to encourage statisticians to demonstrate techniques, especially graphical, for analyzing data and displaying the results of an analysis. Individuals and groups will work with the same set of data and present their analyses at a special session as part of the annual Joint Statistical Meetings in Orlando, Florida on August 13th-17th, 1995. The datasets for 1995 are drawn from two sources, U.S. News & World Report's Guide to Americas Best Colleges and the AAUP (American Association of University Professors) 1994 Salary Survey which appeared in the March-April 1994 issue of Academe. The U.S. News data contains information on tuition, room & board costs, SAT or ACT scores, application/acceptance rates, graduation rate, student/faculty ratio, spending per student, and a number of other variables for 1300+ schools. The AAUP data includes average salary, overall compensation, and number of faculty broken down by full, associate, and assistant professor ranks. The raw data and documentation are contained in the files described below. To obtain any of these files send a message to statlib@lib.stat.cmu.edu of the following form (substituting the file you want for XXXXX) send XXXXX from colleges Available files usnews.doc Documentation for the U.S. News data usnews.data U.S. News data in comma delimited format usnews3.data U.S. News data in fixed column format aaup.doc Documentation for the AAUP salary data aaup.data AAUP salary data in comma delimited format aaup2.data AAUP salary data in fixed column format Two versions of each dataset are provided to accommodate users with different software constraints. The comma delimited versions (USNEWS.DATA and AAUP.DATA) contain information for each college on a separate line with values delimited by commas. The fixed column versions (USNEWS3.DATA and AAUP2.DATA) use 2 or 3 data lines per school and a maximum line length of 80 characters. To participate in the 1995 Data Analysis Exposition you must send an abstract form to the American Statistical Association by February 1st, 1995. Information is available from the ASA Meetings Department by e-mail (meetings@asa.mhs.compuserve.com), phone (703-684-1221), fax (703-684-2037), or surface mail (ASA, 1429 Duke St., Alexandria, VA 22314). Your initial abstract may be fairly general since you may do the bulk of your analysis after the February 1 deadline. You may choose your own path to proceed in analyzing the data or use some of the suggested questions below to get started. ... How well can we model tuition using the other variables? ... How might we cluster colleges into similar comparison groups? ... How can we best display faculty salary structure? ... Can we find a reasonable way to rank the schools? You may work on your own or put together a team. Show off the capabilities of your favorite software package or use the data for a class project and display your students results. You may choose to consider just a subset of schools or examine regional patterns. The main point is to find innovative ways to display the interesting features of the data. Further questions about the 1995 Exposition can be directed to Robin Lock, Mathematics Department, St. Lawrence University, Canton, NY 13617 e-mail rlock@vm.stlawu.edu If you would like to be informed about any subsequent adjustments or error fixes to the 1995 Exposition data, please send an e-mail message to register your interest to rlock@vm.stlawu.edu. Special thanks for providing data for the 1995 Exposition to: Robert Morse, Director of Research for America's Best Colleges at U.S. News & World Report Maryse Eymonerie, Consultant to AAUP. Information about the dataset CLASSTYPE: numeric CLASSINDEX: none specific

17 features

Type (target)nominal4 unique values
0 missing
FICE (ignore)numeric1160 unique values
0 missing
College_name (ignore)nominal1140 unique values
0 missing
Statenominal52 unique values
0 missing
Average_salary-full_professorsnumeric427 unique values
68 missing
Average_salary-associate_professorsnumeric303 unique values
36 missing
Average_salary-assistant_professorsnumeric235 unique values
24 missing
Average_salary-all_ranksnumeric345 unique values
0 missing
Average_compensation-full_professorsnumeric485 unique values
68 missing
Average_compensation-associate_professorsnumeric373 unique values
36 missing
Average_compensation-assistant_professorsnumeric307 unique values
24 missing
Average_compensation-all_ranksnumeric431 unique values
0 missing
Number_of_full_professorsnumeric298 unique values
0 missing
Number_of_associate_professorsnumeric255 unique values
0 missing
Number_of_assistant_professorsnumeric241 unique values
0 missing
Number_of_instructorsnumeric83 unique values
0 missing
Number_of_faculty-all_ranksnumeric495 unique values
0 missing

107 properties

1161
Number of instances (rows) of the dataset.
17
Number of attributes (columns) of the dataset.
4
Number of distinct values of the target attribute (if it is nominal).
256
Number of missing values in the dataset.
87
Number of instances with at least one value missing.
14
Number of numeric attributes.
3
Number of nominal attributes.
0.9
Area Under the ROC Curve achieved by the landmarker weka.classifiers.bayes.NaiveBayes
5.25
First quartile of entropy among attributes.
2.6
Third quartile of skewness among attributes of the numeric type.
0.67
Kappa coefficient achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.51
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .0001
4
Mean kurtosis among attributes of the numeric type.
0.24
Error rate achieved by the landmarker weka.classifiers.bayes.NaiveBayes
0.21
First quartile of kurtosis among attributes of the numeric type.
131.64
Third quartile of standard deviation of attributes of the numeric type.
0.86
Area Under the ROC Curve achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.76
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0.23
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .0001
335.78
Mean of means among attributes of the numeric type.
0.61
Kappa coefficient achieved by the landmarker weka.classifiers.bayes.NaiveBayes
83.74
First quartile of means among attributes of the numeric type.
0.79
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.19
Error rate achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.29
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0.61
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .0001
0.11
Average mutual information between the nominal attributes and the target attribute.
0
Number of binary attributes.
0.11
First quartile of mutual information between the nominal attributes and the target attribute.
0.34
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.67
Kappa coefficient achieved by the landmarker weka.classifiers.bayes.NaiveBayes -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.51
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 3
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .001
45.54
An estimate of the amount of irrelevant information in the attributes regarding the class. Equals (MeanAttributeEntropy - MeanMutualInformation) divided by MeanMutualInformation.
0.43
First quartile of skewness among attributes of the numeric type.
0.41
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 1
0.86
Area Under the ROC Curve achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
33.94
Standard deviation of the number of distinct values among attributes of the nominal type.
0.23
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .001
28
Average number of distinct values among the attributes of the nominal type.
72.17
First quartile of standard deviation of attributes of the numeric type.
0.79
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 2
0.19
Error rate achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.7
Area Under the ROC Curve achieved by the landmarker weka.classifiers.lazy.IBk
0.61
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .001
1.39
Mean skewness among attributes of the numeric type.
5.25
Second quartile (Median) of entropy among attributes.
0.34
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 2
0.67
Kappa coefficient achieved by the landmarker weka.classifiers.lazy.IBk -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.35
Error rate achieved by the landmarker weka.classifiers.lazy.IBk
53.14
Percentage of instances belonging to the most frequent class.
109.26
Mean standard deviation of attributes of the numeric type.
0.54
Second quartile (Median) of kurtosis among attributes of the numeric type.
0.41
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 2
1.43
Entropy of the target attribute values.
0.4
Kappa coefficient achieved by the landmarker weka.classifiers.lazy.IBk
617
Number of instances belonging to the most frequent class.
5.25
Minimal entropy among attributes.
416.4
Second quartile (Median) of means among attributes of the numeric type.
0.79
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.82
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.DecisionStump
5.25
Maximum entropy among attributes.
-0.03
Minimum kurtosis among attributes of the numeric type.
0.11
Second quartile (Median) of mutual information between the nominal attributes and the target attribute.
0.34
Error rate achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.29
Error rate achieved by the landmarker weka.classifiers.trees.DecisionStump
15.58
Maximum kurtosis among attributes of the numeric type.
12.74
Minimum of means among attributes of the numeric type.
0.68
Second quartile (Median) of skewness among attributes of the numeric type.
0.41
Kappa coefficient achieved by the landmarker weka.classifiers.trees.REPTree -L 3
0.47
Kappa coefficient achieved by the landmarker weka.classifiers.trees.DecisionStump
653.49
Maximum of means among attributes of the numeric type.
0.11
Minimal mutual information between the nominal attributes and the target attribute.
0
Percentage of binary attributes.
92.29
Second quartile (Median) of standard deviation of attributes of the numeric type.
0.76
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
0.01
Number of attributes divided by the number of instances.
0.11
Maximum mutual information between the nominal attributes and the target attribute.
4
The minimal number of distinct values among attributes of the nominal type.
7.49
Percentage of instances having missing values.
5.25
Third quartile of entropy among attributes.
0.29
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
12.72
Number of attributes needed to optimally describe the class (under the assumption of independence among attributes). Equals ClassEntropy divided by MeanMutualInformation.
52
The maximum number of distinct values among attributes of the nominal type.
0.34
Minimum skewness among attributes of the numeric type.
1.3
Percentage of missing values.
8.49
Third quartile of kurtosis among attributes of the numeric type.
0.51
Average class difference between consecutive instances.
0.51
Kappa coefficient achieved by the landmarker weka.classifiers.trees.RandomTree -depth 1
0.89
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.J48 -C .00001
3.37
Maximum skewness among attributes of the numeric type.
19.51
Minimum standard deviation of attributes of the numeric type.
82.35
Percentage of numeric attributes.
523.97
Third quartile of means among attributes of the numeric type.
0.86
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.76
Area Under the ROC Curve achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.23
Error rate achieved by the landmarker weka.classifiers.trees.J48 -C .00001
314.09
Maximum standard deviation of attributes of the numeric type.
0.09
Percentage of instances belonging to the least frequent class.
17.65
Percentage of nominal attributes.
0.11
Third quartile of mutual information between the nominal attributes and the target attribute.
0.19
Error rate achieved by the landmarker weka.classifiers.trees.DecisionStump -E "weka.attributeSelection.CfsSubsetEval -P 1 -E 1" -S "weka.attributeSelection.BestFirst -D 1 -N 5" -W
0.29
Error rate achieved by the landmarker weka.classifiers.trees.RandomTree -depth 2
0.61
Kappa coefficient achieved by the landmarker weka.classifiers.trees.J48 -C .00001
5.25
Average entropy of the attributes.
1
Number of instances belonging to the least frequent class.

6 tasks

32 runs - estimation_procedure: 10-fold Crossvalidation - evaluation_measure: predictive_accuracy - target_feature: Type
0 runs - estimation_procedure: 10 times 10-fold Crossvalidation - evaluation_measure: predictive_accuracy - target_feature: Type
0 runs - estimation_procedure: Interleaved Test then Train - target_feature: Type
0 runs - estimation_procedure: 50 times Clustering
0 runs - estimation_procedure: 50 times Clustering
Define a new task