Download JAR files tagged by prediction with all dependencies

Implements StackingC (more efficient version of stacking). For more information, see A.K. Seewald: How to Make Stacking Better and Faster While Also Taking Care of an Unknown Weakness. In: Nineteenth International Conference on Machine Learning, 554-561, 2002. Note: requires meta classifier to be a numeric prediction scheme

Maximum Likelihood Rule Ensembles (MLRules) is a new rule induction algorithm for solving classification problems via probability estimation. The ensemble is built using boosting, by greedily minimizing the negative loglikelihood which results in estimating the class conditional probability distribution. The main advantage of decision rules is their simplicity and comprehensibility: they are logical statements of the form "if condition then decision", which is probably the easiest form of model to interpret. On the other hand, by exploiting a powerful statistical technique to induce the rules, the final ensemble has very high prediction accuracy. Fork of the original code located at: http://www.cs.put.poznan.pl/wkotlowski/software-mlrules.html

A simple meta-classifier that uses a clusterer for classification. For cluster algorithms that use a fixed number of clusterers, like SimpleKMeans, the user has to make sure that the number of clusters to generate are the same as the number of class labels in the dataset in order to obtain a useful model. Note: at prediction time, a missing value is returned if no cluster is found for the instance. The code is based on the 'clusters to classes' functionality of the weka.clusterers.ClusterEvaluation class by Mark Hall.

Performs one-class classification on a dataset. Classifier reduces the class being classified to just a single class, and learns the datawithout using any information from other classes. The testing stage will classify as 'target'or 'outlier' - so in order to calculate the outlier pass rate the dataset must contain informationfrom more than one class. Also, the output varies depending on whether the label 'outlier' exists in the instances usedto build the classifier. If so, then 'outlier' will be predicted, if not, then the label willbe considered missing when the prediction does not favour the target class. The 'outlier' classwill not be used to build the model if there are instances of this class in the dataset. It cansimply be used as a flag, you do not need to relabel any classes. For more information, see: Kathryn Hempstalk, Eibe Frank, Ian H. Witten: One-Class Classification by Combining Density and Class Probability Estimation. In: Proceedings of the 12th European Conference on Principles and Practice of Knowledge Discovery in Databases and 19th European Conference on Machine Learning, ECMLPKDD2008, Berlin, 505--519, 2008.

Class for building pace regression linear models and using them for prediction. Under regularity conditions, pace regression is provably optimal when the number of coefficients tends to infinity. It consists of a group of estimators that are either overall optimal or optimal under certain conditions. The current work of the pace regression theory, and therefore also this implementation, do not handle: - missing values - non-binary nominal attributes - the case that n - k is small where n is the number of instances and k is the number of coefficients (the threshold used in this implmentation is 20) For more information see: Wang, Y (2000). A new approach to fitting linear models in high dimensional spaces. Hamilton, New Zealand. Wang, Y., Witten, I. H.: Modeling for optimal probability prediction. In: Proceedings of the Nineteenth International Conference in Machine Learning, Sydney, Australia, 650-657, 2002.

This package currently contains classes for training multilayer perceptrons with one hidden layer, where the number of hidden units is user specified. MLPClassifier can be used for classification problems and MLPRegressor is the corresponding class for numeric prediction tasks. The former has as many output units as there are classes, the latter only one output unit. Both minimise a penalised squared error with a quadratic penalty on the (non-bias) weights, i.e., they implement "weight decay", where this penalised error is averaged over all training instances. The size of the penalty can be determined by the user by modifying the "ridge" parameter to control overfitting. The sum of squared weights is multiplied by this parameter before added to the squared error. Both classes use BFGS optimisation by default to find parameters that correspond to a local minimum of the error function. but optionally conjugated gradient descent is available, which can be faster for problems with many parameters. Logistic functions are used as the activation functions for all units apart from the output unit in MLPRegressor, which employs the identity function. Input attributes are standardised to zero mean and unit variance. MLPRegressor also rescales the target attribute (i.e., "class") using standardisation. All network parameters are initialised with small normally distributed random values.