> (chi^2-learn-dtree mushroom-data1 mushroom-names) Training data have mixed classification; splitting on some attribute. Splitting on cap-shape yields: (k c b x f) (((e 1)) ((p 2)) ((p 1) (e 6)) ((p 26) (e 59)) ((p 40) (e 65))) The information on this split is: (0 0 .5916727785823275 .8883630233845602 .9587118829771318) Splitting on cap-surface yields: (y s f) (((p 31) (e 62)) ((p 17) (e 15)) ((p 21) (e 54))) The information on this split is: (.9182958340544896 .9971803988942642 .8554508105601307) Splitting on cap-color yields: (c p b w n g y e) (((e 1)) ((e 1)) ((p 6)) ((p 9) (e 12)) ((p 2) (e 45)) ((p 26) (e 37)) ((e 13) (p 26)) ((e 22))) The information on this split is: (0 0 0 .9852281360342516 .25387844029816153 .977895682231035 .9182958340544896 0) Splitting on bruises? yields: (f t) (((p 50) (e 32)) ((p 19) (e 99))) The information on this split is: (.9649567669505688 .6367368786558287) Splitting on odor yields: (c p l a f n) (((p 1)) ((p 5)) ((e 14)) ((e 11)) ((p 58)) ((p 5) (e 106))) The information on this split is: (0 0 0 0 0 .26496353609662915) Splitting on gill-attachment yields: (f) (((p 69) (e 131))) The information on this split is: (.9295231861792264) Splitting on gill-spacing yields: (w c) (((p 4) (e 28)) ((p 65) (e 103))) The information on this split is: (.5435644431995964 .9627730052387542) Splitting on gill-size yields: (n b) (((p 9) (e 11)) ((p 60) (e 120))) The information on this split is: (.9927744539878084 .9182958340544896) Splitting on gill-color yields: (y r k u g w n h p) (((p 2)) ((p 1)) ((p 1) (e 10)) ((e 16)) ((p 15) (e 4)) ((p 5) (e 32)) ((p 4) (e 30)) ((e 8) (p 21)) ((p 20) (e 31))) The information on this split is: (0 0 .4394969869215134 0 .7424875695421236 .5713549744279548 .5225593745369408 .8497511372532974 .966177351388351) Splitting on stalk-shape yields: (e t) (((p 58) (e 34)) ((p 11) (e 97))) The information on this split is: (.9503376699710269 .4748384883008606) Splitting on stalk-root yields: (r c e b) (((e 10)) ((p 2) (e 13)) ((p 5) (e 29)) ((p 62) (e 79))) The information on this split is: (0 .5665095065529053 .6024308020404453 .9894885839592376) Splitting on stalk-surface-above-ring yields: (y f k s) (((p 2)) ((p 4) (e 11)) ((p 47)) ((p 16) (e 120))) The information on this split is: (0 .8366407419411673 0 .5225593745369408) Splitting on stalk-surface-below-ring yields: (y f k s) (((p 2) (e 10)) ((p 8) (e 13)) ((p 47)) ((p 12) (e 108))) The information on this split is: (.6500224216483541 .9587118829771318 0 .4689955935892812) Splitting on stalk-color-above-ring yields: (y g b w n p) (((p 2)) ((e 21)) ((p 13)) ((p 20) (e 78)) ((p 15)) ((p 19) (e 32))) The information on this split is: (0 0 0 .7300166301457939 0 .9526092095121206) Splitting on stalk-color-below-ring yields: (y b g n w p) (((p 2)) ((p 13)) ((e 22)) ((e 3) (p 15)) ((p 20) (e 73)) ((e 33) (p 19))) The information on this split is: (0 0 0 .6500224216483541 .75103241652751 .9470622432515329) Splitting on veil-type yields: (p) (((p 69) (e 131))) The information on this split is: (.9295231861792264) Splitting on veil-color yields: (y w) (((p 2)) ((p 67) (e 131))) The information on this split is: (0 .9232637465337675) Splitting on ring-number yields: (t o) (((p 2) (e 2)) ((p 67) (e 129))) The information on this split is: (1. .9265654986546441) Splitting on ring-type yields: (e l p) (((p 2) (e 26)) ((p 47)) ((p 20) (e 105))) The information on this split is: (.37123232664087563 0 .6343095546405662) Splitting on spore-print-color yields: (w r n h k) (((p 3) (e 5)) ((p 2)) ((p 4) (e 49)) ((p 58)) ((p 2) (e 77))) The information on this split is: (.9544340029249649 0 .3860189005698934 0 .17033057560105985) Splitting on population yields: (c n s a v y) (((p 3)) ((e 9)) ((p 12) (e 23)) ((e 11)) ((e 50) (p 28)) ((p 26) (e 38))) The information on this split is: (0 0 .9275265884316759 0 .9418285354475157 .9744894033980523) Splitting on habitat yields: (l m u p g d) (((p 3) (e 3)) ((p 1) (e 7)) ((p 10) (e 6)) ((e 6) (p 15)) ((p 25) (e 35)) ((p 15) (e 74))) The information on this split is: (1. .5435644431995964 .9544340029249649 .863120568566631 .9798687566511527 .6543524423071836) The information required after splitting on cap-shape is .9015865707518138. The information required after splitting on cap-surface is .9073504806184689. The information required after splitting on cap-color is .6502152152970659. The information required after splitting on bruises? is .771307032856672. The information required after splitting on odor is .1470547625336292. The information required after splitting on gill-attachment is .9295231861792264. The information required after splitting on gill-spacing is .8956996353124889. The information required after splitting on gill-size is .9257436960478215. The information required after splitting on gill-color is .6588335568333942. The information required after splitting on stalk-shape is .693568111869137. The information required after splitting on stalk-root is .842490901029606. The information required after splitting on stalk-surface-above-ring is .4180884303307073. The information required after splitting on stalk-surface-below-ring is .42106344916506877. The information required after splitting on stalk-color-above-ring is .6006234971970297. The information required after splitting on stalk-color-below-ring is .6539682748790425. The information required after splitting on veil-type is .9295231861792264. The information required after splitting on veil-color is .9140311090684299. The information required after splitting on ring-number is .9280341886815512. The information required after splitting on ring-type is .4484159973800765. The information required after splitting on spore-print-color is .207752946130439. The information required after splitting on population is .8414668908874512. The information required after splitting on habitat is .8038724214835198. The attribute with greatest information gain is odor. The attribute odor was chosen as the best. Checking statistical significance... Examining split for value "n" There are 111 examples in this split. For the classification p, 38.3 are expected. (A fraction of .34) There were actually 5. for a difference of -33.3 This contributes 28.95 to Q. For the classification e, 72.7 are expected. (A fraction of .66) There were actually 106. for a difference of 33.3 This contributes 15.25 to Q. Examining split for value "f" There are 58 examples in this split. For the classification p, 20.01 are expected. (A fraction of .34) There were actually 58. for a difference of 37.99 This contributes 72.13 to Q. For the classification e, 37.99 are expected. (A fraction of .66) There were actually 0 for a difference of -37.99 This contributes 37.99 to Q. Examining split for value "a" There are 11 examples in this split. For the classification p, 3.8 are expected. (A fraction of .34) There were actually 0 for a difference of -3.8 This contributes 3.8 to Q. For the classification e, 7.2 are expected. (A fraction of .66) There were actually 11. for a difference of 3.8 This contributes 2. to Q. Examining split for value "l" There are 14 examples in this split. For the classification p, 4.83 are expected. (A fraction of .34) There were actually 0 for a difference of -4.83 This contributes 4.83 to Q. For the classification e, 9.17 are expected. (A fraction of .66) There were actually 14. for a difference of 4.83 This contributes 2.54 to Q. Examining split for value "p" There are 5 examples in this split. For the classification p, 1.72 are expected. (A fraction of .34) There were actually 5. for a difference of 3.28 This contributes 6.22 to Q. For the classification e, 3.28 are expected. (A fraction of .66) There were actually 0 for a difference of -3.28 This contributes 3.28 to Q. Examining split for value "c" There are 1 examples in this split. For the classification p, .34 are expected. (A fraction of .34) There were actually 1. for a difference of .66 This contributes 1.24 to Q. For the classification e, .66 are expected. (A fraction of .66) There were actually 0 for a difference of -.66 This contributes .66 to Q. Evaluating statistical significance with dof=5 and Q=178.87 Probability value is 9.325856423944565e-37 ==> statistically significant, i.e. the theory that this attribute is irrelevant is deemed to be FALSE (probability <= .001) Splitting on this attribute. Now working on the split for the odor="c" with examples:All data have the same classification of: p. Now working on the split for the odor="p" with examples:All data have the same classification of: p. Now working on the split for the odor="l" with examples:All data have the same classification of: e. Now working on the split for the odor="a" with examples:All data have the same classification of: e. Now working on the split for the odor="f" with examples:All data have the same classification of: p. Now working on the split for the odor="n" with examples:Training data have mixed classification; splitting on some attribute. Splitting on cap-shape yields: (b k c x f) (((p 1)) ((e 1)) ((p 2)) ((e 47)) ((p 2) (e 58))) The information on this split is: (0 0 0 0 .21084230031853213) Splitting on cap-surface yields: (y s f) (((p 3) (e 44)) ((p 2) (e 10)) ((e 52))) The information on this split is: (.3424637724817959 .6500224216483541 0) Splitting on cap-color yields: (p b c y e n w g) (((e 1)) ((p 2)) ((e 1)) ((p 2)) ((e 22)) ((e 39)) ((p 1) (e 6)) ((e 37))) The information on this split is: (0 0 0 0 0 0 .5916727785823275 0) Splitting on bruises? yields: (f t) (((p 2) (e 32)) ((p 3) (e 74))) The information on this split is: (.3227569588973983 .23750814382859292) Splitting on gill-attachment yields: (f) (((p 5) (e 106))) The information on this split is: (.26496353609662915) Splitting on gill-spacing yields: (w c) (((p 3) (e 26)) ((p 2) (e 80))) The information on this split is: (.4798320236161285 .1654270339962668) Splitting on gill-size yields: (n b) (((p 3) (e 9)) ((p 2) (e 97))) The information on this split is: (.8112781244591328 .1425733302598987) Splitting on gill-color yields: (r y u h w k n p) (((p 1)) ((p 2)) ((e 16)) ((e 8)) ((p 2) (e 23)) ((e 8)) ((e 25)) ((e 26))) The information on this split is: (0 0 0 0 .4021791902022729 0 0 0) Splitting on stalk-shape yields: (e t) (((p 5) (e 11)) ((e 95))) The information on this split is: (.8960382325345575 0) Splitting on stalk-root yields: (c e b) (((p 2)) ((e 29)) ((p 3) (e 77))) The information on this split is: (0 0 .2307099556914764) Splitting on stalk-surface-above-ring yields: (y f s) (((p 2)) ((e 11)) ((p 3) (e 95))) The information on this split is: (0 0 .19745300438456018) Splitting on stalk-surface-below-ring yields: (y f s) (((p 2)) ((e 13)) ((p 3) (e 93))) The information on this split is: (0 0 .20062232431271465) Splitting on stalk-color-above-ring yields: (y g w p) (((p 2)) ((e 21)) ((p 3) (e 53)) ((e 32))) The information on this split is: (0 0 .30137864359308586 0) Splitting on stalk-color-below-ring yields: (y n p w g) (((p 2)) ((e 3)) ((e 33)) ((p 3) (e 48)) ((e 22))) The information on this split is: (0 0 0 .3227569588973983 0) Splitting on veil-type yields: (p) (((p 5) (e 106))) The information on this split is: (.26496353609662915) Splitting on veil-color yields: (y w) (((p 2)) ((p 3) (e 106))) The information on this split is: (0 .1818131716956182) Splitting on ring-number yields: (t o) (((e 2) (p 2)) ((p 3) (e 104))) The information on this split is: (1. .1844518631409876) Splitting on ring-type yields: (e p) (((p 2) (e 26)) ((p 3) (e 80))) The information on this split is: (.37123232664087563 .22432696034508792) Splitting on spore-print-color yields: (r w n k) (((p 2)) ((p 3) (e 5)) ((e 38)) ((e 63))) The information on this split is: (0 .9544340029249649 0 0) Splitting on population yields: (c a v s y) (((p 3)) ((e 11)) ((p 2) (e 48)) ((e 12)) ((e 35))) The information on this split is: (0 0 .24229218908241482 0 0) Splitting on habitat yields: (m l p u g d) (((p 1)) ((e 3) (p 3)) ((e 2)) ((e 6)) ((p 1) (e 23)) ((e 72))) The information on this split is: (0 1. 0 0 .24988229283318547 0) The information required after splitting on cap-shape is .11396881098299035. The information required after splitting on cap-surface is .21527987717499691. The information required after splitting on cap-color is .03731269774843506. The information required after splitting on bruises? is .2636203934893081. The information required after splitting on gill-attachment is .26496353609662915. The information required after splitting on gill-spacing is .2475688781311856. The information required after splitting on gill-size is .21486574044359968. The information required after splitting on gill-color is .0905808986942056. The information required after splitting on stalk-shape is .1291586641491254. The information required after splitting on stalk-root is .1662774455434064. The information required after splitting on stalk-surface-above-ring is .174327877744927. The information required after splitting on stalk-surface-below-ring is .17351119940559107. The information required after splitting on stalk-color-above-ring is .15204688325416943. The information required after splitting on stalk-color-below-ring is .1482937378717776. The information required after splitting on veil-type is .26496353609662915. The information required after splitting on veil-color is .17853725869209353. The information required after splitting on ring-number is .21384098518996103. The information required after splitting on ring-type is .2613841698611425. The information required after splitting on spore-print-color is .06878803624684432. The information required after splitting on population is .10914062571279946. The information required after splitting on habitat is .10808265790987795. The attribute with greatest information gain is cap-color. The attribute cap-color was chosen as the best. Checking statistical significance... Examining split for value "g" There are 37 examples in this split. For the classification p, 1.67 are expected. (A fraction of .05) There were actually 0 for a difference of -1.67 This contributes 1.67 to Q. For the classification e, 35.33 are expected. (A fraction of .95) There were actually 37. for a difference of 1.67 This contributes .08 to Q. Examining split for value "w" There are 7 examples in this split. For the classification p, .32 are expected. (A fraction of .05) There were actually 1. for a difference of .68 This contributes 1.49 to Q. For the classification e, 6.68 are expected. (A fraction of .95) There were actually 6. for a difference of -.68 This contributes .07 to Q. Examining split for value "n" There are 39 examples in this split. For the classification p, 1.76 are expected. (A fraction of .05) There were actually 0 for a difference of -1.76 This contributes 1.76 to Q. For the classification e, 37.24 are expected. (A fraction of .95) There were actually 39. for a difference of 1.76 This contributes .08 to Q. Examining split for value "e" There are 22 examples in this split. For the classification p, .99 are expected. (A fraction of .05) There were actually 0 for a difference of -.99 This contributes .99 to Q. For the classification e, 21.01 are expected. (A fraction of .95) There were actually 22. for a difference of .99 This contributes .05 to Q. Examining split for value "y" There are 2 examples in this split. For the classification p, .09 are expected. (A fraction of .05) There were actually 2. for a difference of 1.91 This contributes 40.49 to Q. For the classification e, 1.91 are expected. (A fraction of .95) There were actually 0 for a difference of -1.91 This contributes 1.91 to Q. Examining split for value "c" There are 1 examples in this split. For the classification p, .05 are expected. (A fraction of .05) There were actually 0 for a difference of -.05 This contributes .05 to Q. For the classification e, .95 are expected. (A fraction of .95) There were actually 1. for a difference of .05 This contributes 0 to Q. Examining split for value "b" There are 2 examples in this split. For the classification p, .09 are expected. (A fraction of .05) There were actually 2. for a difference of 1.91 This contributes 40.49 to Q. For the classification e, 1.91 are expected. (A fraction of .95) There were actually 0 for a difference of -1.91 This contributes 1.91 to Q. Examining split for value "p" There are 1 examples in this split. For the classification p, .05 are expected. (A fraction of .05) There were actually 0 for a difference of -.05 This contributes .05 to Q. For the classification e, .95 are expected. (A fraction of .95) There were actually 1. for a difference of .05 This contributes 0 to Q. Evaluating statistical significance with dof=7 and Q=91.07 Probability value is 7.444925120753661e-17 ==> statistically significant, i.e. the theory that this attribute is irrelevant is deemed to be FALSE (probability <= .001) Splitting on this attribute. Now working on the split for the cap-color="p" with examples:All data have the same classification of: e. Now working on the split for the cap-color="b" with examples:All data have the same classification of: p. Now working on the split for the cap-color="c" with examples:All data have the same classification of: e. Now working on the split for the cap-color="y" with examples:All data have the same classification of: p. Now working on the split for the cap-color="e" with examples:All data have the same classification of: e. Now working on the split for the cap-color="n" with examples:All data have the same classification of: e. Now working on the split for the cap-color="w" with examples:Training data have mixed classification; splitting on some attribute. Splitting on cap-shape yields: (f c x) (((e 3)) ((p 1)) ((e 3))) The information on this split is: (0 0 0) Splitting on cap-surface yields: (s y f) (((e 3)) ((p 1)) ((e 3))) The information on this split is: (0 0 0) Splitting on bruises? yields: (t f) (((p 1)) ((e 6))) The information on this split is: (0 0) Splitting on gill-attachment yields: (f) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on gill-spacing yields: (w) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on gill-size yields: (n b) (((p 1)) ((e 6))) The information on this split is: (0 0) Splitting on gill-color yields: (n h w k) (((e 2)) ((e 2)) ((p 1)) ((e 2))) The information on this split is: (0 0 0 0) Splitting on stalk-shape yields: (e t) (((p 1)) ((e 6))) The information on this split is: (0 0) Splitting on stalk-root yields: (b e) (((p 1)) ((e 6))) The information on this split is: (0 0) Splitting on stalk-surface-above-ring yields: (f s) (((e 3)) ((p 1) (e 3))) The information on this split is: (0 .8112781244591328) Splitting on stalk-surface-below-ring yields: (f s) (((e 3)) ((p 1) (e 3))) The information on this split is: (0 .8112781244591328) Splitting on stalk-color-above-ring yields: (w) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on stalk-color-below-ring yields: (w) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on veil-type yields: (p) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on veil-color yields: (w) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on ring-number yields: (o) (((p 1) (e 6))) The information on this split is: (.5916727785823275) Splitting on ring-type yields: (p e) (((p 1)) ((e 6))) The information on this split is: (0 0) Splitting on spore-print-color yields: (n w k) (((e 1)) ((p 1)) ((e 5))) The information on this split is: (0 0 0) Splitting on population yields: (a c s) (((e 1)) ((p 1)) ((e 5))) The information on this split is: (0 0 0) Splitting on habitat yields: (l g) (((p 1)) ((e 6))) The information on this split is: (0 0) The information required after splitting on cap-shape is 0. The information required after splitting on cap-surface is 0. The information required after splitting on bruises? is 0. The information required after splitting on gill-attachment is .5916727785823275. The information required after splitting on gill-spacing is .5916727785823275. The information required after splitting on gill-size is 0. The information required after splitting on gill-color is 0. The information required after splitting on stalk-shape is 0. The information required after splitting on stalk-root is 0. The information required after splitting on stalk-surface-above-ring is .46358749969093305. The information required after splitting on stalk-surface-below-ring is .46358749969093305. The information required after splitting on stalk-color-above-ring is .5916727785823275. The information required after splitting on stalk-color-below-ring is .5916727785823275. The information required after splitting on veil-type is .5916727785823275. The information required after splitting on veil-color is .5916727785823275. The information required after splitting on ring-number is .5916727785823275. The information required after splitting on ring-type is 0. The information required after splitting on spore-print-color is 0. The information required after splitting on population is 0. The information required after splitting on habitat is 0. The attribute with greatest information gain is cap-shape. The attribute cap-shape was chosen as the best. Checking statistical significance... Examining split for value "x" There are 3 examples in this split. For the classification p, .43 are expected. (A fraction of .14) There were actually 0 for a difference of -.43 This contributes .43 to Q. For the classification e, 2.57 are expected. (A fraction of .86) There were actually 3. for a difference of .43 This contributes .07 to Q. Examining split for value "c" There are 1 examples in this split. For the classification p, .14 are expected. (A fraction of .14) There were actually 1. for a difference of .86 This contributes 5.14 to Q. For the classification e, .86 are expected. (A fraction of .86) There were actually 0 for a difference of -.86 This contributes .86 to Q. Examining split for value "f" There are 3 examples in this split. For the classification p, .43 are expected. (A fraction of .14) There were actually 0 for a difference of -.43 This contributes .43 to Q. For the classification e, 2.57 are expected. (A fraction of .86) There were actually 3. for a difference of .43 This contributes .07 to Q. Evaluating statistical significance with dof=2 and Q=7. Probability value is .0301973834223185 ==> NOT statistically significant, i.e. the theory that this attribute is irrelevant is deemed to be TRUE (probability > .001) Recursing to pick majority classification. No more attributes to split upon Taking the majority which is: e Now working on the split for the cap-color="g" with examples:All data have the same classification of: e. ;Value 1: (odor (c p) (p p) (l e) (a e) (f p) (n (cap-color (p e) (b p) (c e) (y p) (e e) (n e) (w e) (g e))))