The training _structure of hmm

Source: Internet
Author: User

I think, for most of the time, especially when using hmm in the processing of video data, it is mainly to use the Baumwelch algorithm for learning and training application, that is, the third problem of HMM solution. According to the sample training three sets of parameters in the Hmm.

In the training process, the required samples are labeled well, so, of course, there is no problem. Can use unsupervised also can realize training study, just listen to 52NLP said, unsupervised effect definitely bad. No more detailed analysis, I wonder what is wrong. And in the HMM training learning process, in fact, a bit similar to the implementation of the EM algorithm, initialization of the choice is very important, then what need to pay attention to it. This also did not find the answer.

In the previous process of watching video data, the part that involves the Hmm, is first uses the classification algorithm to realize the classification, then to the classification good sample, namely already has the annotation sample, uses the HMM modelling, the study obtains the hmm the parameter, this and uses the GMM to model is somewhat similar. However, the difference is that there is a coordinate on the state transition time.

However, I am more concerned about the application of hmm in unsupervised training learning. Because the model structure of the non supervision.

There are a lot of problems, probably also I yangaoshoudi, not to say, "Paper on the end feel shallow, absolutely know this matter to preach" well.

Look at the < algorithm yesterday, see the data structure a chapter, think that write well, it is written to my actual situation. said that the structure of the data selection is very important, not to say that is a small part of the program, and sometimes the data structure choice, the program has achieved more than half, can improve the operation of the program and operational efficiency. In fact, no matter what program, especially the more complex program, before programming, always try to want to need how many variables, these amount of what structure. This is very important, and I think as the ability to improve the programming and understanding of the algorithm itself, the choice of programming data structure will become more and more important, good data structure make the program readable and reliable. and bad data structure, will make the program is very difficult to read, and due to poor structure, so that the relationship between data is very confusing, which also brought difficulty for the reliability of the program. Think about it, very messy a large group of data structures, the relationship between the data alone will be difficult to smooth, do not say to use and call.

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.