I am aware that Scipy has a few ICA algorithms, like FastICA, but it can only be used if the mixed signal observations are perfectly in sync.
My application is recording audio (speech) using microphones into mono audio files. So FastICA would not work.
In my research, a few others algorithms I have came across are: Jade, AMUSE and DUET. However, I'm not sure to what extent Python has support for these algorithms. I would prefer to stay in the Python programming language if possible.
Let me just add, I highly value ease of interface, built-in functionality of the Python library, as well as computational efficiency. With that in mind, can someone with experience in using Scipy or other relevant Python libraries suggest a suitable alternative?
I have Anaconda 4.0, and am running Python 3.5 -- just let me know what I should import.
Thank you for reading
答案 0 :(得分：2)
您可以通过安装python中提供的pyemma工具包来运行代码。
conda config --add channels omnia
conda install pyemma
您可以参考此网站获取更多帮助：http://emma-project.org/latest/generated/MSM_BPTI.html