/// \brief Batched convolution operation, with optional window dilation and stride.
///
/// Convolution takes two inputs:
///
/// 1. <i>(the idata batch)</i> a tensor of shape \f$(N,C_\textit{in},d_1,\dots,d_n)\f$ where \f$n > 0\f$, every \f$d_i > 0\f$, and where \f$N\f$ is the batch size
/// (number of data items) and \f$C_\textit{in} > 0\f$ is the number of input channels (sometimes called features); and
/// 2. <i>(the filters)</i> a tensor of shape \f$(C_\textit{out},C_\textit{in},d^f_1,\dots,d^f_n)\f$, where \f$C_\textit{out} > 0\f$ is the number of output channels
/// (sometimes called features) and \f$(d^f_1,\dots,d^f_n)\f$ are the filter dimensions. It is required that for all \f$i\f$, \f$0 < l_i(d^f_i - 1) + 1 \le (d_i - 1)*g_i + 1\f$.
/// (See below for the definition of the window dilation \f$l_i\f$ and the data dilation \f$t_i\f$);
///
/// and five optional parameters:
///
/// 3. <i>(the window movement strides)</i> a vector of positive integers \f$(s_1,\dots,s_n)\f$ (default is all ones),
/// 4. <i>(the window dilation strides)</i> a vector of positive integers \f$(l_1,\dots,l_n)\f$ (default is all ones),
/// 5. <i>(the padding below)</i> a vector of (possibly negative) integers \f$(p_1,\dots,p_n)\f$ (default is all zeros),
/// 6. <i>(the padding above)</i> a vector of (possibly negative) integers \f$(q_1,\dots,q_n)\f$ (default is all zeros), and
/// 7. <i>(the data dilation strides)</i> a vector of non-negative integers \f$(q_1,\dots,q_n)\f$ (default is all ones).
///
/// The output has the shape \f$(N,C_\textit{out},d'_1,\dots,d'_n)\f$, where \f$d'_n = \lceil \frac{(d_i - 1) * t_i + 1 + p_i + q_i - l_i(d^f_i - 1)}{s_i} \rceil\f$.
///
/// Given an input data batch tensor \f$T_\textit{in}\f$, first define the <i>transformed input tensor</i> \f$T_\textit{trans}\f$, with shape \f$(N,C_\textit{in},(d_1 - 1)*t_1+1+p_1+q_1,\dots,(d_n - 1)*t_n+1+p_n+q_n)\f$, as follows:
///
/// \f[
/// T_\textit{trans}[a,c,i_1,\dots,i_n] = T[a,c,(i_1 - p_1)/t_1,\dots,(i_n - p_n)/t_n] \text{ if for all }k, t_k evenly divides (i_k - p_k) and p_k \le i_k \lt p_k + (d_k - 1)*t_k + 1, \text{ else } 0
/// \f]
///
/// then, given an input filter tensor \f$T_\textit{filt}\f$, the output tensor \f$T_\textit{out}\f$ is defined by the equation.
/// \brief Constructs a batched convolution operation with no padding or data dilation (i.e., padding above and below are 0 everywhere, and all data dilation strides are 1).
///
/// \param data_batch The node producing the input data batch tensor.
/// \param filters The node producing the filters tensor.
/// \param window_movement_strides The window movement strides.
/// \param window_dilation_strides The window dilation strides.
/// \param data_batch The node producing the input data batch tensor.<br>
/// `[N, C_IN, D1, ... Df]`
/// \param filters The node producing the filters tensor.<br>
/// `[C_OUT, C_IN, F1, ... Ff]`
/// \param window_movement_strides The window movement strides.<br>
/// `[f]`
/// \param window_dilation_strides The window dilation strides.<br>
/// \brief Constructs a batched convolution operation with no window dilation, padding, or data dilation (i.e., padding above and below are 0 everywhere, and all window/data dilation strides are 1).
///
/// \param data_batch The node producing the input data batch tensor.
/// \param filters The node producing the filters tensor.
/// \param window_movement_strides The window movement strides.
/// \param data_batch The node producing the input data batch tensor.<br>
/// `[N, C_IN, D1, ... Df]`
/// \param filters The node producing the filters tensor.<br>
/// `[C_OUT, C_IN, F1, ... Ff]`
/// \param window_movement_strides The window movement strides.<br>
/// \brief Constructs a batched convolution operation with no window dilation or movement stride (i.e., padding above and below are 0 everywhere, and all window/data dilation strides and window movement strides are 1).
///
/// \param data_batch The node producing the input data batch tensor.
/// \param filters The node producing the filters tensor.
/// \param data_batch The node producing the input data batch tensor.<br>
/// `[N, C_IN, D1, ... Df]`
/// \param filters The node producing the filters tensor.<br>
