First, thank you for this brilliant synthesis of deep clustering algorithms!
I wonder, however, if slightly more emphasis could be given to the data types commonly seen in the clinical epidemiological/register-based context, such as categorical (including binary) data, as well as mixed categorical and continuous data, also often seen in such studies. The use of deep learning in this field is very promising, but it is not as well-compiled as deep clustering on naturally high-dimensional data such as images. In the literature, one may find recent papers focusing on this issue, e.g., the DeepTLF framework (https://doi.org/10.1007/s41060-022-00350-z) and recent breakthroughs should be highlighted. It would be very beneficial if a section could be added for these types of clustering methods in this repo.
First, thank you for this brilliant synthesis of deep clustering algorithms!
I wonder, however, if slightly more emphasis could be given to the data types commonly seen in the clinical epidemiological/register-based context, such as categorical (including binary) data, as well as mixed categorical and continuous data, also often seen in such studies. The use of deep learning in this field is very promising, but it is not as well-compiled as deep clustering on naturally high-dimensional data such as images. In the literature, one may find recent papers focusing on this issue, e.g., the DeepTLF framework (https://doi.org/10.1007/s41060-022-00350-z) and recent breakthroughs should be highlighted. It would be very beneficial if a section could be added for these types of clustering methods in this repo.
Thanks again!
Daniil