Artificial Intelligence on the Sinclair QL: Make Your Micro by Keith Brain, Steven Brain

By Keith Brain, Steven Brain

Show description

Read or Download Artificial Intelligence on the Sinclair QL: Make Your Micro Think PDF

Best computers books

Application and Theory of Petri Nets 1993: 14th International Conference Chicago, Illinois, USA, June 21–25, 1993 Proceedings

This quantity includes the court cases of the 14th foreign convention onApplication and thought of Petri Nets. the purpose of the Petri web meetings is to create a discussion board for discussing development within the program and thought of Petri nets. in most cases, the meetings have 150-200 individuals, one 3rd of whom come from undefined, whereas the remainder are from universities and learn institutes.

Digital Image processing.6th.ed

The 6th version has been revised and prolonged. the complete textbook is now basically partitioned into uncomplicated and complicated fabric for you to do something about the ever-increasing box of electronic snapshot processing. during this method, you could first paintings your approach throughout the easy rules of electronic picture processing with no getting crushed through the wealth of the cloth after which expand your reviews to chose themes of curiosity.

Extra info for Artificial Intelligence on the Sinclair QL: Make Your Micro Think

Example text

IF DE% was >=0, but the computer was wrong (and selected CAR). then the weighting variable. WT%. is set to —1. If DE% was <0, but the computer was wrong (and selected BICYCLE), then WT% is set to +1. 13000 DEFins PROCedure BICYCLE IF INS="Y" THEN WT%=0 : UPDATE_RULE 13010 13020 WT%=-1 ^ UPDATE RULE 13030 END DEFine BICYCLE 14000 DEFine PROCedure CAR IF IN$="Y" THEN WT0 : UPDATE_RULE 14010 14020 W^%=1 ^ UPDATE_RULE 14030 END DEFine CAR The effect of the weighting variable takes place in the UPDATE_RULE PROCedure in which we modify the values in the rule array, RU(N), pulling them down when they are too high.

It is easier to avoid errors if these are entered into DATA in matching pairs and READ in turn into the array. Stan a new program with these lines which SETUP the array. ) In the same wa y . the first three characters IN$(1 TO 3) could be checked against 'YOU' and replaced when necessary by 'I': 80 IF IN #(1 TO 3)= "YOL' THEN PRINT "I" (4 TO) & IN$ If you try that out with a series of sentences you will see that it works OK until you type something like: YOU ARE TIRED which comes back as the rather unintelligent: I ARE TIRED We could get around this by checking for the phrases 'I AM' and 'YOU ARE' as well as 'I and `YOU' on their own.

If both of these are true then the rules are updated again b y subtracting the correct feature values. FV(N), to bias in favour of the correct answer. 180 UPDATE RULEB 19000 DEFine PROCedure LJPD4TE_RULEB 19010 FOR N=1 TO OB% IF DE(N)>=DE% AND N{}CR% THEN 19020 19030 FOR M=1 TO FE% 19040 RU(M,N)=((RU( ^ ,N)*^Ρ3) (F ^ (M)#3))/3 19050 END FOR M 19060 ELSE NEXT N 19070 END IF 19080 END FOR N Then the correct feature values. FV(N). are added to the rule array for the correct object to bias in the opposite direction.

Download PDF sample

Rated 4.28 of 5 – based on 31 votes