【正文】
ake more papers.2. How to find a valuable problems.I mentioned that the scientific society is a chain constructed of different research groups. every research group use result of other research groups. try to find out which group of research teams or individual researcher use your papers and then try to ply their needs. just think science as a bussiness. this business needs marketting to improve. then do this marketing before selection of a topic to research. read papers of people who use your works even speak with them or do joint work with them to find out their needs. then i repeat again that dont just read papers in your topic read paper of people who use your papers to see their need. then READ MORE PAPERS.3. Just think science as a business. Marketing is important factor in success.Must researchers just think about what they can do interesting with their skills and laboratory facilities . but it is not true. just think about people who read your papers and want to use it. how you can help them. what they need. imagine your self as a virtual reader to see how a reader can use your work.4. Develop a suitable model not a very general and plex modelsdont just try to develop a very plicated model which can solver more general problems then befire. most general method is already exist. people can model their systems with modlling constructing atoms and their interaction with quantum mechanics but it is a really impossible one!. plex models have two important disadvantages as below1. first disadvantage of more plex models it is harder to describe a system. plex models often need more variables to describe a system and finding these values are not so easy in practice. then people need to estimate these variables and these result to their accuracy bee low as like as less accurate models.2. plex models need more time to learning, programming, and even in putation time.then they never would use plex models if simpler models can result similar try to develop a model which is logical. note that your model is always different from real model which are in use in practice then try a model which have logical simplicfication. always remember very simple formuala for error of the result. error is the result is greatest error of error producers. then if you ignore one source of e