Home >> All >> org >> joone

•	org.joone.dte.* (16)	•	org.joone.engine.* (98)	•	org.joone.exception.* (2)
•	org.joone.inspection.* (5)	•	org.joone.io.* (33)	•	org.joone.log.* (4)
•	org.joone.net.* (12)	•	org.joone.samples.* (22)	•	org.joone.script.* (5)
•	org.joone.structure.* (1)	•	org.joone.util.* (46)

Package Samples:

• org.joone.dte.selection
• org.joone.dte.factory
• org.joone.dte.control
• org.joone.dte
• org.joone.exception
• org.joone.io
• org.joone.engine.weights
• org.joone.engine.listeners
• org.joone.engine.learning
• org.joone.engine.extenders
• org.joone.engine
• org.joone.util
• org.joone.structure
• org.joone.log
• org.joone.samples.engine.multipleInputs
• org.joone.samples.engine.parity
• org.joone.samples.engine.scripting
• org.joone.samples.engine.timeseries
• org.joone.samples.engine.validation
• org.joone.samples.engine.xml

Classes:

• JDBCInputSynapse: The JDBCInputSynapse provides support for data extraction from a database. To use this synapse the user should ensure a JDBC Type 4 Driver is in the class path. It is possible to use other JDBC driver types though you will have to refer to the vendors documentation, it may require extra software insallation and this may limit your distributtion to certain Operating Systems. The properties required by this JDBCInputSynapse Plugin are the following Database Driver Name - e.g sun.jdbc.odbc.JdbcOdbcDriver Database URL - e.g jdbc:mysql://localhost/MyDb?user=myuser&password=mypass SQLquery - e.g select ...
• JDBCOutputSynapse: The JDBCOutputSynapse provides support for data input to a database. To use this synapse the user should ensure a JDBC Type 4 Driver is in the class path. It is possible to use other JDBC driver types though you will have to refer to the vendors documentation, it may require extra software insallation and this may limit your distributtion to certain Operating Systems. The properties required by this JDBCOutputSynapse Plugin are the following Database Driver Name - e.g sun.jdbc.odbc.JdbcOdbcDriver Database URL - e.g jdbc:mysql://localhost/MyDb?user=myuser&password=mypass SQLAmendment - e.g "INSERT ...
• KohonenSynapse: This is an unsupervised Kohonen Synapse which is a Self Organising Map. This KohonenSynapse works in conjunction with the next layer which should implement a SOM strategy such as a GuassianLayer or WTALayer (Winner Takes All). This synapse should connect to one of these layers, without a SOM Strategy in the next layer this component will not function correctly. This KohonenSynapse takes a pattern from the previous layer, calculates the distance between the input vector and the weights and passes this on to the next layer. In the learning phase the next layer should calculate the distance fall off ...
• UnNormalizerOutputPlugIn: UnNormalizes the input data within a predefined range. To enable the UnNormalizer to find the min and max within the input data specify zero values for setInDataMin and setInDataMax. To set user defined values for the input data max and min in a serie then specify non-zero values for setInDataMin and setInDataMax. The PlugIn supports two modes - Buffered and UnBuffered. Buffered Mode If the StreamOutputSynapse that this PlugIn is attached to is in buffered mode then the PlugIn can either search for input data min/max values if getInDataMin()==0 and getInDataMax()==0 or if either of these methods ...
• YahooFinanceInputSynapse: The YahooFinanceInputSynapse provides support for financial data input from financial markets. The synapse contacts YahooFinance services and downloads historical data for the chosen symbol and date range. Finally the data is presented to the network in reverse date order i.e oldest first. This synapse provides the following info .. Open as column 1 High as column 2 Low as column 3 Close as column 4. Volume as column 5. Adj.Close as column 6. For the particular stock symbol. Developer Notes : This YahooFinanceInputSynapse uses the following format to extract stock financial information from the ...
• TaskFactory: This interface must be implemented by any class that is capable to create tasks to elaborate (i.e. neural nets to train) starting from external parameters written in some formalism or language (e.g. XML) After the TaskFactory is created, the DTE calls on it the initialize() method, and then, for each neural network it needs to elaborate, the getNextTask() is called, until it returns null. The developer has two possibilities, based on the creation mechanism he wants to implement: 1 - In order to avoid to occupy many memory, the initialize method doesn't create neural networks, as they are created ...
• Layer: The Layer object is the basic element forming the neural net. Primarily it consists of a number of neurons that apply a transfer function to the sum of a number of input patterns and convey the result to the output pattern. The input patterns are received from connected input listeners and the transformed results are passed to connected output listeners. The component also handles learning by accepting patterns of error gradients from output listeners, applying a reverse (inverse) transfer function and passing the result to the input listeners. Layers execute their own Threads to perform the perform ...
• GaussianLayer: This layer implements the Gaussian Neighborhood SOM strategy. It receives the euclidean distances between the input vector and weights and calculates the distance fall off between the winning node and all other nodes. These are passed back allowing the previous synapse to adjust it's weights. The distance fall off is calculated according to a Gaussian distribution from the winning node. This layer uses implemtations of SpatialMap in order to calculate these distances. Currently this layer uses the GaussianSpatialMap which calculates the Gaussian distance for all nodes in the SOM map. Future maps ...
• DeltaNormPlugIn: This plugin calculates the Delta Normalization on a time series. the Delta Normalization technique permits to feed a neural network with the delta values instead of the absolute values of the series, permitting in this manner to avoid the problems correlated with both ascending and descending trends. The normalization is obtained by dividing the delta value by the dynamic range (named Probability Volatility Windows) calculated on the given period. This plugin is very useful for financial predictions when used in combination with the MinMaxExtractorPlugin. To learn more on this technique read the ...
• NeuralNetAttributes: This class represents a descriptor of a neural network. It stores some parameters of the neural network useful to manage the training and the validation error of the net without having it loaded in memory. Created for the distributed environment to store the 'state' of the training of a net without the necessity to load in memory all the nets belonging to a generation, it can be extended to store more parameters for any other future use. WARNING: This class must be compatible with the SOAP-Serialization mechanism, hence add ONLY public parameters having: - public getter/setter methods - a basic ...
• FanInBasedWeightInitializer: The weights are uniformly distributed (that is randomly) within the range [LB/F_i, UB/F_i] . LB and UB stand for lower bound and upper bound , which is a certain number. Here the bounds will be by default -2.4 and 2.4 as described in Neural Networks - A Comprehensive Foundation, Haykin , chapter 6.7 Some Hints for Making the Back-Propagation Algorithm Perform Better . F_i is the fan-in, i.e. the total number of inputs) of neuron i. There is also an option to use instead of F_i the square root of F_i , which is also used in some cases.
• MovingAveragePlugIn: Changes the specified input serie data so that it becomes a moving average of itself. This plugin operates on specified serie/s of data in a vertical fashion. For example if the serie to be converted contained the following data .... 5 15 5 and the requested moving average was set at 2 then the serie would become 0 10 12.5 Any data prior to the moving average spec is set at 0 as there is not enough data to calculate the actual moving average. The data is NOT normalised. To normalise the data use a NormalizerConverterPlugIn .
• ExtendableLearner: Learners that extend this class are forced to implement certain functions, a so-called skeleton. The good thing is, because learners extend this class certain plug-ins can be added. For example, plug ins that change the objective function, or the delta-update rule. Still learners that do not fit into this skeleton have to opportunity to implement Learner directly (or extend AbstractLearner), but it won't be able to use the extra plug-ins (unless it is build in the learner by the programmer itself). Basically, this class is the BasicLearner, but by adding extenders it can provide totally different ...
• NetErrorManager: This class should be used when ever a critical error occurs that would impact on the training or running of the network. Joone classes should construct a new NetErrorManager when an error occurs. By doing so this will stop and reset the network so the user can perform corrective action and re-start. E.g new NetErrorManager(monitor,"An error has occurred.!"); The constructor of this class calls the monitors fireNetStoppedError event method which propogates the event to all the net listeners. This in turn stops and resets the network to allow correction and continuation.
• FahlmanTeacherSynapse: This class extends the normal Teacher synapse and implements the Fahlman 40-20-40 criterion (the values can be changed). This teacher makes only sense in case of binary outputs. In case of the default values (40-20-40) and considering [0,1] binary outputs the criterion is fullfilled if for all patterns the output is within [0, 0.4] in case where the desired output is 0 and the output is within [0.6, 1] in case where the desired output is 1. More obout this criterion can be found at http://citeseer.ist.psu.edu/fahlman88empirical.html .
• RmseCondition: This condition verifies if the actual population has reached the maxRmse specified as parameter. if validation = true, the validation rmse is checked, otherwise the control is done on the training rmse (the default). if mean = true, the condition is verified only if the average rmse of the population is below maxRmse, otherwise the control is done on the fittest neural network. As this class extends the MaxCyclesCondition, also a max number of iterations can be specified, in order to stop the DTE if the maxRmse is not reached after that number of cycles.
• ScriptValidationSample: This example shows how to use the joone's scripting engine to to check the training level of the net using a validation data source. In this example we will learn to use the following objects: - org.joone.util.LearningSwitch - org.joone.util.MacroPlugin - org.joone.util.NormalizerPlugIn This program shows how to build the same neural net contained into the org/joone/samples/editor/scripting/ValidationSample.ser file using only java code and the core engine's API. Open that net in the GUI editor to see the architecture of the net built in this example.
• SimpleValidationSample: This example shows how to check the training level of the net using a validation data source. In this example we will learn to use the following objects: - org.joone.util.LearningSwitch - org.joone.net.NeuralNetValidator - org.joone.util.NormalizerPlugIn This program shows how to build the same kind of neural net as that contained into the org/joone/samples/editor/scripting/ValidationSample.ser file using only java code and the core engine's API. Open that net in the GUI editor to see the architecture of the net built in this example.
• MultipleValidationSample: This example shows how to check the training level of a neural network using a validation data source. The training and the validation phases of the created network is executed many times, showing for each one the resulting RMSE. This program shows how to build the same kind of neural net as that contained into the org/joone/samples/editor/scripting/ValidationSample.ser file using only java code and the core engine's API. Open that net in the GUI editor to see the architecture of the net built in this example.
• Nakayama: This class performs the method of optimizing activation functions as described in: K.Nakayama and Y.Kimura, "Optimization of activation functions in multilayer neural network applied to pattern classification" , Proc. IEEE ICNN'94 Florida, pp.431-436, June 1994. This techniques probably fails whenever the NeuralNet.join() method is called because this optimization technique stops the network to perform the optimization, use a NeuralNetListener instead.
• FIRFilter: Element of a connection representing a FIR filter (Finite Impulse Response). The DelaySynapse object implements a delayed full synapse where each connection is implemented with a FIRFilter object. In this connection is implemented the temporal backpropagation algorithm by Eric A. Wan, as in 'Time Series Prediction by Using a Connectionist Network with Internal Delay Lines' in Time Series Prediction. Forecasting the Future and Understanding the Past, by A.Weigend and N.Gershenfeld. Addison-Wesley, 1994.
• ComparingElement: This interface describes an element that can compare the output of the layer to which it is connected, with another input derivating from a StreamInputSynapse named 'desired'. To elaborate the result of the comparison, attach to its output a whatever component implementing the OutputPatternListener interface. (use addResultSynapse to do it). Its main purpose is to describe the interface of a component used to teach the neural network, but it can be used whenever it's necessary to compare two patterns.
• ColumnSelectorPlugIn: Certain plug-ins change the number of columns during their conversion, for example the ToBinaryPlugin increases the number of columns. This plug in can be used to select certain columns from all available columns at that moment. It just uses an advanced column selector as StreamInputSynapse and all plug-ins do, but this plug-in does not work on certain columns as most plug-ins do, but it selects columns like the StreamInputSynapse does.
• SimulatedAnnealingExtender: Simulated annealing (SA) refers to the process in which random or thermal noise in a system is systematically decreased over time so as to enhance the system's response. Basically the change of weights and biases in SA is defined as: dW = dw + (n)(r)(2^-kt), where dw is the weight / bias change produced by standard back propagation, n is a constant controlling the initial intensity of the noise, k is the decay constant,t is the generation counter and r is a random number.
• NeuralNet: This object represents a container of a neural network, giving to the developer the possibility to manage a neural network as a whole. Thanks to it, a neural network can be saved and restored using an unique writeObject and readObject command, without be worried about its internal composition. Not only this, because using a NeuralNet object, we can also easily transport a neural network on remote machines and runnit there, writing only few and generalized java code.

Home | Contact Us | Privacy Policy | Terms of Service