This chapter mainly organizes protocol calculation, segmentation, sequence comparison and index extraction in TensorFlow, which is convenient for future use
1. Statute calculation
The operation of the statute calculation will have the function of dimensionality reduction. In all the reduce_xxx series of operation functions, the dimensionality reduction is performed by means of XXX. Each function has the axis parameter, that is, along a certain direction, use the xxx method to perform the input Tensor Dimensionality reduction.
Note: The default value of axis is None, which means that input_tensor is reduced to 0 dimension, which is a number.
For two-dimensional input_tensor : axis=0, calculated by column; axis=1, calculated by row.
Second, segmentation
The segmentation operation is an uncommon operation in TensorFlow, and it is occasionally used in complex network models.
3. Sequence comparison and index extraction
Operations on sequences and arrays:
You can pay attention to the official account of me and my friends~~~ Here are some python technical information that my friends and I update from time to time! ! You can also leave a message to discuss technical issues, I hope you can support, pay attention to it, click on the small advertisement, thank you~~
