About Me

Hi ,


Myself Sakthi Dasan (alias shakthydoss) a graduate in B. Tech Information Technology, I preferred to have good discussions on all topics which comes under the taxonomy of computer science   I am more interested in challenging jobs and research opportunities.

contact me : shakthydoss@gamil.com


17 Responses to “About Me”

  1. Dr.N.C.GHATAK Says:

    I was a Medical Librarian & resigned from the post.
    Now I am writing on Medical topics & Librarianship & doing medical services to the needed persons of the society.
    I found that you are highly learned person on the IT technical field. I do not know about this IT. You can
    give me your knowledge. However we will be in touch.

    With the good wishes to you.
    Thanks with regards to you.

    • shakthydoss Says:

      Dear Dr.N.C.GHATAK

      Thank you for the kind wishes, I am really willing to share my knowledge if that is going to benefit others and it really motivate me also .

      We could definitely be in touch; my email id shakthydoss@gamil.com

      And I think we both have already met before; when I come to West Bengal for competition INFOCOM you stepped into my stall, hope you can recall that [?]

      • Drncghatak Ghatak Says:

        Dear Sakti,
        How are you? I am OK. I am now working at the Government Medical College as Librarian.
        I would like to go to Antarctica. You can accompany me. Please arrange to go there.
        Give me all the information regarding this Antarctica tour.
        With the best wishes to you.

  2. Abhishek R Nath Says:

    I am Abhishek R. Nath. I am doing my M.Tech in Computer Science in Amrita Vishwa Vidyapeetham. As part of the case study that was given to us, we are working in Hidden Markov Models.
    Our topic is Extracting the Names of Genes and Gene Products with a Hidden Markov Model. We have to predict genes and protein words from a given biological document. We completed almost all part of our case study. We designed an HMM with three states- Protein ,Gene, Unknown. So the transition matrix is a 3X3 matrix. But we have a confusion in selecting the dimension for the observation matrix. We have designed a 3X3 matrix for observation probability distribution. But we are getting a wrong output. Our model is predicting non protein words as protein, gene etc. We are putting some random values in observation matrix since we dont know how actually the values of observation matrix are calculated. Our guide is also saying the same that our calculation of values for observation matrix may be wrong.

    While searching its answer through the internet, I came across your HMM tool for address validation. I downloaded it and executed it. Then I thought you must know how to calculate the values for the observation matrix. We need your help. So please help us. Please tell us how to calculate the probabilities. If you want more details please reply to me, I will send the details.

    • shakthydoss Says:

      Dear Abhishek ,

      Observation matrix

      row corresponds to your observation and observation can be a target-words or keywords.
      columns corresponds to your state.
      In your case Observation matrix size can be nX3

      calculating the value :
      Assume a matrix like this
      S1 S2 S3
      A 2/14 5/14 0
      B 1/14 3/14 1/14
      C 11/14 1/14 4/14
      D 3/14 6/14 0/14
      E 0 7/14 2/14

      Now look at first row of fist column in above matrix its value is 2/14 , 1.e the the target words A which belong to state S1has occurred 2 times in the entire document and it should be divided by the total no of target words in the entire document.

      simply say keyword Frequency / total no of Keywords
      I tied this formula for my HMM tool .

      • Abhishek R Nath Says:

        But I have a doubt. How the target word A can appear in both the state S1 and S2 at the same time? At a time a single word can appear in only one state, right? And also what is n in nX3 means, is it the total no. of words in the document?

  3. shakthydoss Says:

    To Abhishek

    It is not necessary that target words have to occur in only one state. It depends on the application what you develop. In your case target word have to take only one state so make the frequency count accountable for that particular state and make zero for other stares.

    eg :

    2/14 0 0
    0 4/14 0
    6/14 0 0
    0 0 4/14

    Then ‘n’ is the no of target words in your document

  4. Mari Says:

    Generar la matrix count

    • shakthydoss Says:

      I cant understand your question , sorry.

  5. Mustafa adel Says:

    can I have your face book acoount ?

  6. Sathish Says:

    I want to know about Key in cryptography.
    I cant understand what Key is about.. I want a example from u.. thanks in advance

    • shakthydoss Says:

      Hi Sathish ,

      Key is nothing but a piece of information that helps to crack you cipher text.

      Example : –
      Consider General bank lock and key system
      in which unique lock can only be opened by its specific key.

      And depending upon the cryptography technology this key system various.

  7. Roshan Chaudhary Says:

    hi there,
    my name is roshan chaudhary and i am really interested in your blog and i have also mailed you regarding my issues. hope you check the email and reply me your views.

    Roshan Chaudhary

    • shakthydoss Says:

      Dear Roshan ,

      could not get your issue ……
      can you revert back the mail again.

  8. divya Says:

    Im persuing my M.E (computer science engineering) and im doing project in the area web mining(i.e) personalization.can u share ur idea about preprocessing the web documents. Thank u.

    with regards.,

  9. A.Subramani Says:

    A.Subramani 9003161140

    the letters in the uyirmeiyazhthukal tamil you mentioned uyiryzhuthukal

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: