Preskoči na glavno vsebino

Artificial intelligence


          Artificial intelligence (AI) is a relatively young branch of science that stirs the imagination of many. Even movie directors from hollywood are not exceptions. Development in AI area is very fast and there is no indication that this will change soon. I still remember my first contact with learning devices. This happend at the end of the last millennium when I realized neural networks (NN). They have immediately attracted my attention, because such devices were not known till then.
   

      NN are made along the lines of mammalian brain. During the learning NN extract the essence from the data. After the learning we can ask NN questions. It gives us the right answers even to questions that during learning did not participate.  NN learns autonomously and therefore may exceed the teacher's (poeple's) knowledge. Here are some important achievements of artificial intelligence:


  • A couple of years ago the computer Deep Blue became the world champion of chess.



  • In the year 2011 a computer beat the two best human competitors on the US show Jeopardy.



  • Google already have a personal vehicle, which drives automatically on normal public roads and follows all traffic regulations.






  • Honda's famous robot ASIMO


  • Converting speech into written word and speech translation in a foreign language just experiencing a scientific breakthrough. After decades of effort machines have recorded and translated speech almost as good as the people.






And now a few words about my research work in the field of AI. I made my first research on the field of speech recognition. This is a complicated problem. I tried to find out, which NN would be suitable for the recognition of different speech voices (phonemes). I built a suitable NN and tought it to recognize syllables with the accuracy of about 95%.

My second research was on the field of robotics. I built a NN that moves the legs of a six-legged, ant-like robot. The robot can successfully overcome uneven ground and small steps.








One of the biggest problems on which I stumbled upon during the use of NN, was problematic learning. Like humans NN have problems with learning. The most known process of learning for the NN is  ‘‘Backpropagation’’. This learning process has some disadvantages. I dug into that and found a couple of ideas for improvement:

  • I have named my first improvement ‘‘Bipropagation’’. With it, I achieved a more than ten-times acceleration of learning. The learning is also more reliable. Bipropagation also solves very complex problems, where the Backpropagation method completely fails. It is a kind of a Deep learning. Demo App
  • The second improvement is called ‘‘Border Pairs Method’’ (BPM). This improvement first eliminates unimportant learning data. During the learning process, it finds the optimal construction of the NN and it doesn’t come to overfitting. After the learning, the NN answers correct all learning questions.

In essence, we could say, that the BPM method isn’t learning to memorize, but learning to understand. On top of that, the BPM method qualitatively allows us  removing of noise in the learning data, finding features and clustering. The BPM method combines the good qualities of the SVM method and Backpropagation.


In July 2014 a book entitled Advances in Machine Learning Research (Nova Publishers, New York) was published. It describes the mentioned algorithms (Bipropagation and BPM).

I’m looking forward to your response here or in our Artificial Intelligence community on the Google+.
الذكاء الاصطناعي

intelligence artificielle

intelligenza artificiale

人工知能

künstliche Intelligenz

искусственный интеллект

inteligencia artificial

Komentarji

  1. I am psychotherapist and I am very much excited about the computer modelling of understanding that is actually the essence of my practice. I wonder about your insights on that issue and definitely would like to know more.

    OdgovoriIzbriši
  2. I am psychotherapist and I am very much interested in machine understanding. Actually "understanding" is my core professional activity. I would appreciate your insights into that issue. The question is "how to model the process of deep understanding of one human being by another?".

    OdgovoriIzbriši
    Odgovori
    1. Dear Denis!

      Human understanding is for me new but interesting topic.
      What do you think about a chat by gmail about it.

      Bojan

      Izbriši
  3. Let's chat, it will be a real pleasure. I am mostly free to chat during evenings, starting from 20.00 Kiev time. During daytime too busy in understanding humans :)

    OdgovoriIzbriši
  4. I have just started a Machine Learning course and I find it quite fascinating. I read a lot about AI and I believe it is the future of computing .

    OdgovoriIzbriši
    Odgovori
    1. Sorry for late response. I totally agree with you. Do you like research?

      Izbriši

Objavite komentar

Priljubljene objave iz tega spletnega dnevnika

Bionika, zanimiv poklic prihodnosti

Živimo v času, ko se znanje ustvarja in širi vedno hitreje in temu trendu se pridružujemo tudi izobraževalne ustanove. Ponudba smeri in stopenj izobraževalnih programov se vsako leto povečuje in med novimi programi prevladujejo interdisciplinarni. To ustreza manjšim podjetjem, ki so bolj prožna in odzivna in se zato lahko bolje znajdejo v tem gospodarsko razburkanem času. V malih podjetjih namreč ni prostora za množico ozko usmerjenih strokovnjakov, v prednosti so kadri z interdisciplinarno širino, ki jo omogočata mehatronika in predvsem bionika .      Na višji strokovni šoli na Ptuju smo že pred časom zaznali piš tega novega vetra in začeli prikrojevati naša jadra. Prvi korak v to smer je bil program Mehatronika , ki smo ga razvili pred približno desetletjem in povezuje med seboj več tehniških strok. Predvsem so to elektrotehnika, računalništvo in strojništvo. Trg delovne sile je potrdil našo vizijo saj se naši diplomanti uspešno zaposlujejo tako v Sloveniji, kakor tudi v t

Beyond Backpropagation

Gartner is predicting a very bright near future for the "Machine learning". 2015 was a peak year of inflated expectations, now, in 2016 is following period of disillusionment and in 2017 should be reached the plateau of productivity. Elsewhere this process usually last for 10 years. One kind of the most popular modern "machine learning" is named "Deep Learning" what is another name for neural networks with little bit more layers and perhaps even with a convolution and/or recursion. The learning of this kinds networks was until now usually based on gradient descent, on slow, iterative, non-reliable process named Backpropagation . That kind of learning is very demanding and extensive. On plain computer can last for hours or even many days and is often unsuccessful concluded. Recently are appeared two algorithms that significantly improve this kind of machine learning: " Bipropagation " and " Border pairs method ". Bipropagat