概要
kerasでbackendにsubtractが、無い。
一覧
K.abs Element-wise absolute value.
K.all Bitwise reduction (logical AND).
K.any Bitwise reduction (logical OR).
K.arange Creates a 1D tensor containing a sequence of integers.
K.argmax Returns the index of the maximum value along an axis.
K.argmin Returns the index of the minimum value along an axis.
K.backend Active Keras backend
K.batch_dot Batchwise dot product.
K.batch_flatten Turn a nD tensor into a 2D tensor with same 1st dimension.
K.batch_get_value Returns the value of more than one tensor variable.
K.batch_normalization Applies batch normalization on x given mean, var, beta and gamma.
K.batch_set_value Sets the values of many tensor variables at once.
K.bias_add Adds a bias vector to a tensor.
K.binary_crossentropy Binary crossentropy between an output tensor and a target tensor.
K.cast_to_floatx Cast an array to the default Keras float type.
K.cast Casts a tensor to a different dtype and returns it.
K.categorical_crossentropy Categorical crossentropy between an output tensor and a target tensor.
K.clear_session Destroys the current TF graph and creates a new one.
K.clip Element-wise value clipping.
K.concatenate Concatenates a list of tensors alongside the specified axis.
K.constant Creates a constant tensor.
K.conv1d 1D convolution.
K.conv2d_transpose 2D deconvolution (i.e. transposed convolution).
K.conv2d 2D convolution.
K.conv3d_transpose 3D deconvolution (i.e. transposed convolution).
K.conv3d 3D convolution.
K.cos Computes cos of x element-wise.
K.count_params Returns the static number of elements in a Keras variable or tensor.
K.ctc_batch_cost Runs CTC loss algorithm on each batch element.
K.ctc_decode Decodes the output of a softmax.
K.ctc_label_dense_to_sparse Converts CTC labels from dense to sparse.
K.cumprod Cumulative product of the values in a tensor, alongside the specified axis.
K.cumsum Cumulative sum of the values in a tensor, alongside the specified axis.
K.depthwise_conv2d 2D convolution with separable filters.
K.dot Multiplies 2 tensors (and/or variables) and returns a tensor.
K.dropout Sets entries in x to zero at random, while scaling the entire tensor.
K.dtype Returns the dtype of a Keras tensor or variable, as a string.
K.elu Exponential linear unit.
K.epsilon Fuzz factor used in numeric expressions.
K.equal Element-wise equality between two tensors.
K.eval Evaluates the value of a variable.
K.exp Element-wise exponential.
K.expand_dims Adds a 1-sized dimension at index "axis".
K.eye Instantiate an identity matrix and returns it.
K.flatten Flatten a tensor.
K.floatx Default float type
K.foldl Reduce elems using fn to combine them from left to right.
K.foldr Reduce elems using fn to combine them from right to left.
K.function Instantiates a Keras function
K.gather Retrieves the elements of indices indices in the tensor reference.
K.get_session TF session to be used by the backend.
K.get_uid Get the uid for the default graph.
K.get_value Returns the value of a variable.
K.get_variable_shape Returns the shape of a variable.
K.gradients Returns the gradients of variables w.r.t. loss.
K.greater_equal Element-wise truth value of (x >= y).
K.greater Element-wise truth value of (x > y).
K.hard_sigmoid Segment-wise linear approximation of sigmoid.
K.identity Returns a tensor with the same content as the input tensor.
K.image_data_format Default image data format convention ('channels_first' or 'channels_last').
K.in_test_phase Selects x in test phase, and alt otherwise.
K.in_top_k Returns whether the targets are in the top k predictions.
K.in_train_phase Selects x in train phase, and alt otherwise.
K.int_shape Returns the shape of tensor or variable as a list of int or NULL entries.
K.is_keras_tensor Returns whether x is a Keras tensor.
K.is_placeholder Returns whether x is a placeholder.
K.is_sparse Returns whether a tensor is a sparse tensor.
K.l2_normalize Normalizes a tensor wrt the L2 norm alongside the specified axis.
K.learning_phase Returns the learning phase flag.
K.less_equal Element-wise truth value of (x <= y).
K.less Element-wise truth value of (x < y).
K.local_conv1d Apply 1D conv with un-shared weights.
K.local_conv2d Apply 2D conv with un-shared weights.
K.log Element-wise log.
K.logsumexp Computes log(sum(exp(elements across dimensions of a tensor))).
K.manual_variable_initialization Sets the manual variable initialization flag.
K.map_fn Map the function fn over the elements elems and return the outputs.
K.max Maximum value in a tensor.
K.maximum Element-wise maximum of two tensors.
K.mean Mean of a tensor, alongside the specified axis.
K.min Minimum value in a tensor.
K.minimum Element-wise minimum of two tensors.
K.moving_average_update Compute the moving average of a variable.
K.ndim Returns the number of axes in a tensor, as an integer.
K.normalize_batch_in_training Computes mean and std for batch then apply batch_normalization on batch.
K.not_equal Element-wise inequality between two tensors.
K.one_hot Computes the one-hot representation of an integer tensor.
K.ones_like Instantiates an all-ones variable of the same shape as another tensor.
K.ones Instantiates an all-ones tensor variable and returns it.
K.permute_dimensions Permutes axes in a tensor.
K.placeholder Instantiates a placeholder tensor and returns it.
K.pool2d 2D Pooling.
K.pool3d 3D Pooling.
K.pow Element-wise exponentiation.
K.print_tensor Prints message and the tensor value when evaluated.
K.prod Multiplies the values in a tensor, alongside the specified axis.
K.random_binomial Returns a tensor with random binomial distribution of values.
K.random_normal_variable Instantiates a variable with values drawn from a normal distribution.
K.random_normal Returns a tensor with normal distribution of values.
K.random_uniform_variable Instantiates a variable with values drawn from a uniform distribution.
K.random_uniform Returns a tensor with uniform distribution of values.
K.relu Rectified linear unit.
K.repeat_elements Repeats the elements of a tensor along an axis.
K.repeat Repeats a 2D tensor.
K.reset_uids Reset graph identifiers.
K.reshape Reshapes a tensor to the specified shape.
K.resize_images Resizes the images contained in a 4D tensor.
K.resize_volumes Resizes the volume contained in a 5D tensor.
K.reverse Reverse a tensor along the specified axes.
K.rnn Iterates over the time dimension of a tensor
K.round Element-wise rounding to the closest integer.
K.separable_conv2d 2D convolution with separable filters.
K.set_epsilon Fuzz factor used in numeric expressions.
K.set_floatx Default float type
K.set_learning_phase Sets the learning phase to a fixed value.
K.set_value Sets the value of a variable, from an R array.
K.set_session TF session to be used by the backend.
K.set_image_data_format Default image data format convention ('channels_first' or 'channels_last').
K.shape Returns the symbolic shape of a tensor or variable.
K.sigmoid Element-wise sigmoid.
K.sign Element-wise sign.
K.sin Computes sin of x element-wise.
K.softmax Softmax of a tensor.
K.softplus Softplus of a tensor.
K.softsign Softsign of a tensor.
K.sparse_categorical_crossentropy Categorical crossentropy with integer targets.
K.spatial_2d_padding Pads the 2nd and 3rd dimensions of a 4D tensor.
K.spatial_3d_padding Pads 5D tensor with zeros along the depth, height, width dimensions.
K.sqrt Element-wise square root.
K.square Element-wise square.
K.squeeze Removes a 1-dimension from the tensor at index "axis".
K.stack Stacks a list of rank R tensors into a rank R+1 tensor.
K.std Standard deviation of a tensor, alongside the specified axis.
K.stop_gradient Returns variables but with zero gradient w.r.t. every other variable.
K.sum Sum of the values in a tensor, alongside the specified axis.
K.switch Switches between two operations depending on a scalar value.
K.tanh Element-wise tanh.
K.temporal_padding Pads the middle dimension of a 3D tensor.
K.tile Creates a tensor by tiling x by n.
K.to_dense Converts a sparse tensor into a dense tensor and returns it.
K.transpose Transposes a tensor and returns it.
K.truncated_normal Returns a tensor with truncated random normal distribution of values.
K.update_add Update the value of x by adding increment.
K.update_sub Update the value of x by subtracting decrement.
K.update Update the value of x to new_x.
K.var Variance of a tensor, alongside the specified axis.
K.variable Instantiates a variable and returns it.
K.zeros_like Instantiates an all-zeros variable of the same shape as another tensor.
K.zeros Instantiates an all-zeros variable and returns it.
以上。