it's known that WEKA craches when the input dataset is too big. For this reason you have to run the algorithms from your terminal avoiding the GUI. Because the length of the commands is too big, I developed this programm which takes as input the installation directory of WEKA, the directory of dataset and the output directory. Then a menu with available algorithms will appeared and you have to choose one. Finally when the algorithm has terminated the results is visible in the terminal.
$ git clone https://github.com/sp1thas/WEKACMDs.git && cd WEKACMDs
$ python WekaCommands.py-
Python 2.7
-
termcolor
Installation (run as root):
$ pip install -r requirements.txt- WEKA link
- Your dataset
$ python WekaCommands.py -i <inputfile> -o <outputfile> -w <wekadirectory> -i, --ifile
This is the input dataset
-o, --ofile
This is the output file with classification results
(model is not contained)
-w, --wekadir
Direction with WEKA software
-h,
Prints these options
| Bayes | Availability |
|---|---|
| BayesNet | ✓ |
| NaiveBayes | ✓ |
| NaiveBayesMultinomial | ✓ |
| NaiveBayesMultinomialText | ✓ |
| NaiveBayesUpdateable | ✓ |
| Functions | Availability |
|---|---|
| Logistic | ✓ |
| MultilayerPerceptron | ✓ |
| SimpleLogistic | ✓ |
| SMO | ✓ |
| Lazy | Availability |
|---|---|
| IBk | ✓ |
| KStar | ✓ |
| LWL | ✓ |
| Meta | Availability |
|---|---|
| AdaBoostM1 | - |
| AdditiveRegression | - |
| AttributeSelectedClassifier | - |
| Bagging | ✓ |
| ClassificationViaRegression | - |
| CostSensitiveClassifier | - |
| CVParameterSelection | - |
| FilteredClassifier | - |
| IterativeClassifierOptimizer | - |
| LogitBoost | - |
| MultiClassClassifier | - |
| MultiClassClassifierUpdateable | - |
| MultiScheme | - |
| RandomCommittee | - |
| RandomizableFilteredClassifier | - |
| RandomSubSpace | - |
| RegressionByDiscretization | - |
| Stacking | - |
| Vote | - |
| WeightedistancesHandlerWrapper | - |
| Misc | Availability |
|---|---|
| InputMappedClassifier | ✓ |
| SerializedClaassifier | - |
| Rules | Availability |
|---|---|
| DecisionTable | ✓ |
| JRip | ✓ |
| M5Rules | - |
| OneR | ✓ |
| PART | ✓ |
| ZeroR | ✓ |
| Trees | Availability |
|---|---|
| DecisionTableDecisionStump | ✓ |
| HoeffdingTree | ✓ |
| J48 | ✓ |
| LMT | ✓ |
| M5P | - |
| RandomForest | ✓ |
| RandomTree | ✓ |
| REPTree | ✓ |
| Trees | Availability |
|---|---|
| RBFNetwork | ✓ |
run python script with necessery arguments
press enter to continue
choose algorith
check results
and output file has been generated
- Simakis Panagiotis - Initial work
This project is licensed under the GNU General Public License version 3 - see the LICENSE file for details