# ML final project ## model: SVM > 報告者: 施孟成 --- - total 1692 images - Preprocessing: gray scale => resize to 64*64 - Build 5 classifiers for each attributes. - Validation method: * Holdout validation *HV* * Cross validation *CV* --- ## take HP(生命值) attribute for example --- ### kernel: linear | validation method| HV(train) | HV(test) | CV | | -------- | -------- | -------- | -------- | | Score | .830 | .167 | .162 | --- | validation method| HV | CV | | -------- | -------- | -------- | | confusion matrix | |  | --- ### kernel: poly | validation method| HV(train) | HV(test) | CV | | -------- | -------- | -------- | -------- | | Score | .649 | .191 | .173 | --- | validation method| HV | CV | | -------- | -------- | -------- | | confusion matrix |  |  | --- ### kernel: rbf | validation method| HV(train) | HV(test) | CV | | -------- | -------- | -------- | -------- | | Score | .364 | .177 | .189 | --- | validation method| HV | CV | | -------- | -------- | -------- | | confusion matrix | |  | --- ### kernel: sigmoid | validation method| HV(train) | HV(test) | CV | | -------- | -------- | -------- | -------- | | Score | .188 | .169 | .172 | --- | validation method| HV | CV | | -------- | -------- | -------- | | confusion matrix | |  | --- ## Let's see another attributes... --- ### PA(物理攻擊)  kernel: linear score: .189 --- ### MA(魔法攻擊)  kernel: linear score: .179 --- ### SP(敏捷or速度)  kernel: poly score: .178 --- ### CR(爆擊)  kernel: poly score: .159 ###### tags: `機器學習概論`