(3)在EM(PENE找到一列,记为第j列,它含有最少数目 的死元素“*”,A←人{x (4)如果EM(PENE)删去所有在第j列含有非死元素的行; (5)如果EM(PENE)=⑨,则终止,并返回A否则1+1 并转向(3) 参考文献: 归纳学习一算法,理论,应用。洪加荣著P46-52,P.57-58
(3) 在EM(PE|NE)找到一列，记为第j列，它含有最少数目 的死元素“*”,A←A {xj}. (4) 如果EM(PE|NE) 删去所有在第j列含有非死元素的行； (5) 如果EM(PE|NE)= ,则终止，并返回A;否则i←I+1. 并转向(3); 参考文献： 归纳学习—算法，理论，应用。 洪加荣著 P.46-52, P.57-58.
Training Examples for Enjoy Sport Sky Temp Humid Wind Water Forecst EnjoySpt Sunny Warm Normal Strong Warm Same Yes Sunny Warm High Strong Warm Same Ye Rainy Cold High Strong Warm Change No Sunny Warm High Strong Cool Change Yes
Prototypical Concept Learning Task ● Giver: Instances X: Possible days, each described by the attributes Sky, Air Temp, Humidity, Wind. water, forecast Target function c: Enjoy Sport: X>10,1] Hypotheses H: Conjunctions of literals. E (?,Cold,High,"?,"?,? Training examples D: Positive and negative examples of the target function 〈x1,c(x1)),…(xm,c(xm) Determine: A hypothesis h in H such that h(a)=c(a) for all a in D
表 FIND-S算法 1.将h初始化为H中最特殊假设 2.对每个正例x 对h的每个属性约束a 如果x满足a 那么不做任何处理 否则将h中a替换为x满足的另一个更一般约束 3.输出假设h h←<0,0.0.0.0.0> he<Sunny, Warm, Normal, Strong, Warm, Same> he<Sunny Warm Strong, Warm. Same> he<Sunny, Warm,?, Strong, ?,?
1. 将h初始化为H中最特殊假设 2. 对每个正例x 对h的每个属性约束a 如果x满足a 那么不做任何处理 否则将h中a替换为x满足的另一个更一般约束 3. 输出假设h h←<Ø , Ø , Ø , Ø , Ø , Ø > h←<Sunny,Warm,Normal,Strong,Warm,Same> h←<Sunny,Warm,?,Strong,Warm,Same> h←<Sunny,Warm,?,Strong,?,?> 表 FIND-S算法
The List-Then-Eliminate algorithm 1. Version Space < a list containing every hypothesis in H 2. For each training example, (a, c(a) remove from Version Space any hypothesis h fo widh(x)≠c(x) 3. Output the list of hypotheses in Version space