This app takes the following arguments:
int numSamples (500): number of MNIST training samples to use
int patchSize (10): size of patch edge in pixels to feed into the AHaH tree
int resolution (8): number of features generate=2^resolution.
main(String[]) -
Static method in class com.mancrd.ahah.samples.classifier.mnist.MnistApp
This app takes the following arguments:
int poolSize (~8): pool size in pixels
int patchSize = (~8): size of patch to feed into the AHaH tree
int encoderResolution = (~10): number of features produced per tree=2^depth
int numEncoders = (~3): more trees-->more features-->better performance-->longer run-time.
double unsupervisedConfidenceThreshold (1.0): confidence threshold, above which unsupervised learning will take place
double startAuto (.25): percentage through data when unsupervised learning starts;
double learningRate (.5): classifier learning rate
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
isFunctional (true) : functional or circuit AHaH model
isSweetSpot (true) : does experiment averaging at sweetspot orthogonal value if true.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
frequency (100): frequency of voltage source
timeStep (1E-4): time step of simulation
amplitude (.25): amplitude of voltage source
totalTime (5E-2): total simulation time.
This app takes the following arguments:
int numJoints (9): the number of joints in the robotic arm
int startLevel (0): the level that that game should start at, zero-indexed
int numTargetsPerLevel (100): the number of pills the arm should catch before switching to the next level
int numFibersPerMuscle (20): the number of muscle fibers each muscle has
int bufferLength (1): the buffer length - controls the delay time in which the proprioceptive data feeds back into the motor controller
Random points along a line picked from a Gaussian distribution centered at N blobs are eventually clustered and labeled in an unsupervised manner
by the clusterer
Static1d() -
Constructor for class com.mancrd.ahah.samples.clusterer.visual.oned.Static1d