Optimizable

abstract val optim: Optimizer

optimizer object

abstract def step(data: RDD[Vector]): Optimizable.this.type

Perform a gradient-based optimization step

final def !=(arg0: Any): Boolean

final def ##(): Int

final def ==(arg0: Any): Boolean

final def asInstanceOf[T0]: T0

var batchSize: Option[Int]

Minibatch size for each iteration in the ascent procedure.

def clone(): AnyRef

var converged: Boolean

convergence flag for local optimisation

var convergenceTol: Double

Error tolerance in log-likelihood for the stopping criteria

final def eq(arg0: AnyRef): Boolean

def equals(arg0: Any): Boolean

def finalize(): Unit

def getBatchSize: Option[Int]

final def getClass(): Class[_]

def getConvergenceTol: Double

def getLoss: Double

def getMaxIter: Int

def getOptim: Optimizer

def getSeed: Long

implicit var globalIterCounter: Long

this prevents the seed from repeating every time step() is called which would cause the same samples being taken

def hashCode(): Int

final def isInstanceOf[T0]: Boolean

var lossValue: Double

Current loss value

implicit val matrixOps: ParameterOperations[DenseMatrix[Double]]

var maxIter: Int

Maximum number of iterations allowed

final def ne(arg0: AnyRef): Boolean

final def notify(): Unit

final def notifyAll(): Unit

var regularizer: Option[Regularizer]

Optional regularization term

var seed: Long

random seed for mini-batch sampling

def setBatchSize(n: Option[Int]): Optimizable.this.type

def setBatchSize(n: Int): Optimizable.this.type

def setConvergenceTol(x: Double): Optimizable.this.type

def setMaxIter(m: Int): Optimizable.this.type

def setOptim(optim: Optimizer): Optimizable.this.type

def setRegularizer(r: Option[Regularizer]): Optimizable.this.type

def setRegularizer(r: Regularizer): Optimizable.this.type

def setSeed(s: Long): Optimizable.this.type

def step(data: DStream[Vector]): Optimizable.this.type

Update model parameters using streaming data

def step(data: JavaRDD[Vector]): Unit

Perform a gradient-based optimization step

final def synchronized[T0](arg0: ⇒ T0): T0

def toString(): String

implicit val vectorOps: ParameterOperations[DenseVector[Double]]

Linear Algebra operations necessary for computing updates for the parameters

final def wait(): Unit

final def wait(arg0: Long, arg1: Int): Unit

final def wait(arg0: Long): Unit