knuth

for 45 years and  knuth was my technical author of choice. I have possessed his publishing books for about three years. now i possess all other writings including book on his search in bible. I have read them about 50%. When I was young i used to spend my time solving problems from art of programming series. experimentally i searched technology supremo google for max(a,b) or min(a.b) as a function of arithmetic function of a and b. after many iteration i gave up. I wish google consumes Art of Computer Programming books and other books that we learn from.

thieves are committed for life time in stealing. perfecting the art of stealing.

I was not at all interested in his MIX machine and his programs written in them. Only recently nothing interesting to read i picked up his book on combinatorics. it is the first part of the 4th volume of Art of Computer programming. someway i like writing algorithms for problems of combinatorics. I enjoyed magic square as a child. i came across various strategies involving Magic squares and its extensions to latin squares and graeco-latin squares etc. in design of experiment. this is very interesting topic in determining the perfect combination that yields better result.

suppose we have 6 different age groups. 6 different types of cancer. 6 different treatments. we wish to differentiate in which age group for which cancer type which treatment is preferable.

this book started with graeco latin square and generating them. i learnt this problem was attempted by the famous mathematician euler 250 years back. he was nearly blind them but he used use a number of students of directing them to search for solution. the problem was solved by rcbose parker etc. parker wrote a program following euler's method bringing down search to be within reach of computation. not only for 6 10 14 etc which previously thought by euler having no solution. this solution in 1955 appeared in the first page ny times. never before a mathematical discovery made to first page of a daily newspaper. 

Suppose in addition to 6 cancer type, 6 treatment, 6 different age groups, we have 6 different diets, can we have 6x6x6x6 magic 4D where all the combinations occur once and only once. greco latin crelic. with sufficient number of samples in position of the 4D cube we may be able to determine what is the prefered way to treat a patient with some clue towards making error estimation. use of computer in solving combinatorial problems may have some practical usage in clinical treatment or better yields in farming output.

my frustration is with technology misusage searching voluminous petabytes of garbage search made by google and any number of organization in search of truth. google itself employs 10000 linux human installers. mere size of data collection does not make it book of truth. we must create data intelligently where we may discover gems of truth.

i find my purposeless idle reading clears my blindness poisonous atmosphere created by peddlers of drugs and thieves.

Comments

Popular posts from this blog

non american

প্রান্তরের গান আমার

dinkar