awin/Obsidian-Main
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

vault backup: 2023-11-14 10:25:23

Browse Source
This commit is contained in:
Awin Huang
2023-11-14 10:25:23 +08:00
parent 1d53fb47de
commit 573de783d6
4 changed files with 65 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
.obsidian/workspace.json vendored
View File

@@ -34,9 +34,21 @@
},
"pinned": true
}
},
{
"id": "26dbd5276842579d",
"type": "leaf",
"state": {
"type": "markdown",
"state": {
"file": "04. Programming/Machine Learning.md",
"mode": "source",
"source": true
}
}
}
],
"currentTab": 1
"currentTab": 2
}
],
"direction": "vertical"
@@ -65,7 +77,7 @@
"state": {
"type": "search",
"state": {
"query": "numpy ",
"query": "softma",
"matchingCase": false,
"explainSearch": false,
"collapseAll": false,
@@ -82,7 +94,8 @@
"state": {}
}
}
]
],
"currentTab": 1
}
],
"direction": "horizontal",
@@ -102,7 +115,7 @@
"state": {
"type": "backlink",
"state": {
"file": "00. Inbox/01. TODO.md",
"file": "04. Programming/Machine Learning.md",
"collapseAll": false,
"extraContext": false,
"sortOrder": "alphabetical",
@@ -127,7 +140,7 @@
"state": {
"type": "outline",
"state": {
"file": "00. Inbox/01. TODO.md"
"file": "04. Programming/Machine Learning.md"
}
}
},
@@ -155,12 +168,12 @@
"state": {
"type": "file-properties",
"state": {
"file": "00. Inbox/01. TODO.md"
"file": "04. Programming/Machine Learning.md"
}
}
}
],
"currentTab": 5
"currentTab": 3
},
{
"id": "ae4bf98badbfc7ee",
@@ -198,13 +211,16 @@
"periodic-notes:Open today": false
}
},
"active": "3b4577823cedf427",
"active": "26dbd5276842579d",
"lastOpenFiles": [
"04. Programming/categorical_crossentropy.md",
"05. 資料收集/An example that use categorical_crossentropy and softmax.md",
"04. Programming/Machine Learning.md",
"00. Inbox/01. TODO.md",
"01. 個人/01. Daily/2023-11-13(週一).md",
"05. 資料收集/讀書筆記/20231113 - 戰爭裡的世界史.md",
"05. 資料收集/讀書筆記/20230801 - 蘇格拉底哲學特快車.md",
"01. 個人/01. Daily/2023-11-12(週日).md",
"00. Inbox/01. TODO.md",
"00. Inbox/除草劑.md",
"05. 資料收集/Tool Setup/Software/Visual Studio Code.md",
"01. 個人/01. Daily/2023-11-11(週六).md",
@@ -224,10 +240,7 @@
"01. 個人/01. Daily/2023-10-26(週四).md",
"未命名.canvas",
"04. Programming/Python/matplotlib.md",
"01. 個人/01. Daily/2023/08/2023-08-01(週二).md",
"01. 個人/01. Daily/2023/09",
"01. 個人/01. Daily/2023-10-24(週二).md",
"01. 個人/01. Daily/2023/09/2023-09-04(週一).md",
"01. 個人/01. Daily/2023/08",
"01. 個人/01. Daily/2023/05",
"00. Inbox/My Mindmap.canvas",

14
04. Programming/Machine Learning.md
View File

@@ -1,11 +1,11 @@
## 問題分類
| 問題類型 | 啟動函數 | 損失函數 |
|:-----------:|:--------:|:--------------------------------------:|
| 二元分類 | sigmoid | binary_crossentropy（二元交叉熵） |
| 單標籤多元分類 | softmax | [[categorical_crossentropy]]（分類交叉熵）<br> sparse_categorical_crossentropy |
| 多標籤分類 | sigmoid | binary_crossentropy |
| 回歸求值 | None | mse（均方誤差） |
| 回歸求0~1值 | sigmoid | mse或binary_crossentropy |
| 問題類型 | 啟動函數 | 損失函數 | 範例 |
|:--------------:|:--------:|:------------------------------------------------------------------------------:|:------------------------------------------------------------------:|
| 二元分類 | sigmoid | binary_crossentropy（二元交叉熵） | |
| 單標籤多元分類 | softmax | [[categorical_crossentropy]]（分類交叉熵）<br> sparse_categorical_crossentropy | 範例：[[An example that use categorical_crossentropy and softmax]] |
| 多標籤分類 | sigmoid | binary_crossentropy | |
| 回歸求值 | None | mse（均方誤差） | |
| 回歸求0~1值 | sigmoid | mse或binary_crossentropy | |

3
04. Programming/categorical_crossentropy.md
View File

@@ -1,2 +1,3 @@
- 僅適用於 one-hot 編碼。
- 如果輸出不是 one-hot，而是整數標籤，也就是直接輸出 0、1、2，而不是一個array（`[0, 0, 0, 1, 0]` 之類），那就需要 sparse_categorical_crossentropy。
- 如果輸出不是 one-hot，而是整數標籤，也就是直接輸出 0、1、2，而不是一個array（`[0, 0, 0, 1, 0]` 之類），那就需要 sparse_categorical_crossentropy。
- 範例：[[An example that use categorical_crossentropy and softmax]]

31
05. 資料收集/An example that use categorical_crossentropy and softmax.md Normal file
View File

@@ -0,0 +1,31 @@
An example that use [[categorical_crossentropy]] and softmax
```python
dropRatio = 0.1
model = Sequential()
model.add(Conv1D(filters=32, kernel_size=3, activation='relu', input_shape=(self.DATA_LEN, 1)))
model.add(Dropout(dropRatio))
model.add(MaxPooling1D(pool_size=2))
model.add(Conv1D(filters=64, kernel_size=3, activation='relu'))
model.add(Dropout(dropRatio))
model.add(MaxPooling1D(pool_size=2))
model.add(Conv1D(filters=64, kernel_size=3, activation='relu'))
model.add(Dropout(dropRatio))
model.add(MaxPooling1D(pool_size=2))
model.add(Conv1D(filters=128, kernel_size=3, activation='relu'))
model.add(MaxPooling1D(pool_size=2))
model.add(Dropout(dropRatio))
model.add(Conv1D(filters=256, kernel_size=3, activation='relu'))
model.add(MaxPooling1D(pool_size=2))
model.add(Dropout(dropRatio))
model.add(Flatten())
model.add(Dropout(dropRatio))
model.add(Dense(units=128, activation='relu'))
model.add(Dropout(dropRatio))
model.add(Dense(units=64, activation='relu'))
model.add(Dropout(dropRatio))
model.add(Dense(units=32, activation='relu'))
model.add(Dense(units=len(self.DEVICE_LABEL), activation='softmax'))
model.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate=0.01), metrics=['accuracy'])
```