AD Teaching Wiki:

Results for Exercise Sheet 10 (Naive Bayes)

Please read the instructions below, before adding something to the table!

Add your row to the table below, following the examples already there. Column 2, 3, and 4 = running time of your program for reading the CSV file, training, and prediction, respectively, in seconds (with exactly one digit after the dot). Column 5 = percentage of your predictions that were correct. Column 5 = machine specification as usual (processor frequency with exactly one digit after the dot, amount of RAM as an integer, no secondary details about processor). Column 6 = programming language (Java or C++). Column 7 = A short description of the feature selection improvements you made (if you made any).

Name

Reading time

Training time

Prediction time

Correct %

Processor / RAM

Language

Feature Selection

Florian B

7.3s

14.9s

9.5s

49%

Intel X5560 @ 2.8GHz / 36GB

Java

Florian B

7.9s

15.0s

8.6s

53%

Intel X5560 @ 2.8GHz / 36GB

Java

no stopwords

Christoph S

15.5s

0.4s

0.7s

53%

Core 2 Quad @ 2.8GHz / 4GB

C++

Christoph S

36.6s

1.1s

4.2s

62%

Core 2 Duo @ 1.6GHz / 2GB

C++

no stopwords

Christoph S

34.3s

0.0s

0.0s

21%

Core 2 Duo @ 1.6GHz / 2GB

C++

Just predict "class 0" all the time ;)

Christoph S

34.3s

1.1s

4.1s

67%

Core 2 Duo @ 1.6GHz / 2GB

C++

No stopwords. 1653 training points. Used docs with multiple class labels (counting them).

Matthias H.

24.8s

0.4s

1.7s

42%

Intel i3 @ 1.3GHz / 4GB

Java

Stefan

10.0s

0.1s

43.0s

47%

Intel i5 @ 1.7GHz / 4GB

Java

Anthony

7.0s

6.0s

23.0s

18%

Intel i7 @ 2.10GHz / 8GB

Java

Adrian B.

3.5s

0.1s

0.1s

47%

Intel i7 @ 3.4GHz / 16GB

C++

Adrian B.

3.0s

0.1s

0.1s

55%

Intel i7 @ 3.4GHz / 16GB

C++

no stopwords

Adrian B.

3.0s

0.1s

0.1s

67%

Intel i7 @ 3.4GHz / 16GB

C++

no stopwords. Use every tenth doc for training, no matter how many labels it has.

Aritz B

7.7s

0.1s

5.4s

26%

Intel Core 2 Duo @ 2.4GHz / 6GB

Java

Tobias S

23.9s

1.1s

46.9s

22%

Intel Atom @ 1.6GHz / 2GB

C++

Chris Sch

17.8s

0.2s

16.0s

37%

Core 2 Duo @ 2.5GHz / 4GB

Java

Simon S

18.8s

0.1s

8.9s

27%

Intel X5560 @ 2.8GHz / 36GB

C++

Ander B

8.4s

1.6s

12.8s

26%

Intel Core i5 @ 2.3GHz / 8GB

Java

Ane R

7.6s

1.0s

4.6s

27%

Intel i5-2450M @ 2.5GHz / 8GB

Java

Alves J

7.9s

4.2s

17.1s

29%

Intel Core 2 Duo @ 2.4GHz / 4GB

Java

Tobias F

10.3s

3.4s

24.8s

19%

Phenom II X4 965 @ 3.4GHz / 4GB

Java

Jan M

21.3s

0.3s

1.5s

54%

Intel E5645 @ 2.4GHz / 23GB

C++

Jan M

18.2s

0.3s

1.2s

62%

Intel E5645 @ 2.4GHz / 23GB

C++

no stopwords

Matthias F

10.0s

1.0s

8.3.2s

37%

Intel i7 CPU @ 2.10GHz / 8GB

C++

Andreas H

3.8s

0.03s

1.9s

50%

Intel P8700 @ 2.53GHz / 4GB

C++

word length >= 9

AD Teaching Wiki: InformationRetrievalWS1213/ResultsNaiveBayes (last edited 2013-01-23 13:10:35 by Hannah Bast)