.

.github/workflows/check_links.yml

@@ -0,0 +1,21 @@
+name: Check links
+
+on:
+  push:
+  pull_request:
+  schedule:
+    - cron: "0 0 * * 0" # At 00:00 on Sunday
+
+jobs:
+  check-links:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Check
+        uses: gaurav-nelson/[email protected]
+        with:
+          use-quiet-mode: "yes"
+          use-verbose-mode: "yes"
+          config-file: ".github/workflows/check_links_config.json"

.github/workflows/check_links_config.json

@@ -0,0 +1,25 @@
+{
+  "ignorePatterns": [
+    {
+      "pattern": [
+        "https://www.ollydbg.de/version2.html",
+        "https://www.onlinegdb.com/",
+        "https://whoer.net/",
+        "https://store.nethunter.com/",
+        "https://www.microsoft.com/en-us/edge",
+        "https://www.epicgames.com/",
+      ]
+    }
+  ],
+  "httpHeaders": [
+    {
+      "urls": ["https://github.com/", "https://guides.github.com/", "https://help.github.com/", "https://docs.github.com/"],
+      "headers": {
+        "Accept-Encoding": "zstd, br, gzip, deflate"
+      }
+    }
+  ],
+  "timeout": "20s",
+  "retryOn429": true,
+  "aliveStatusCodes": [200, 206]
+}

.github/workflows/deploy.yml

@@ -0,0 +1,20 @@
+name: Deploy
+
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 1
+
+      - name: Deploy
+        run: |
+          pip install mkdocs-material
+          mkdocs gh-deploy --force

docs/README.md

@@ -0,0 +1,26 @@
+# Notes
+
+示例笔记:
+
+- [常用软件](操作系统/常用软件.md).
+- [Vim](其他/编辑器/Vim.md).
+- [反汇编器](渗透测试/逆向工程/反汇编器.md).
+- [Manjaro 安装](操作系统/Linux/Manjaro/Manjaro_安装.md).
+- [Dishonored 2](<其他/游戏/评价/Dishonored 2.md>).
+
+活跃模块:
+
+- 计算机网络.
+- 数据库开发.
+- [现代推箱子教程](https://shenmian.github.io/sokoban-tutorial/).
+
+## 许可协议
+
+![License](assets/LICENSE.png){ align=right }
+
+该系列文章采用 [CC BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/) 许可协议进行授权.  
+部分图片来源于互联网, 版权归原作者所有.
+
+---
+
+[![Give up GitHub](https://sfconservancy.org/img/GiveUpGitHub.png){ width=20% style="display: block; margin: 0 auto" }](https://sfconservancy.org/GiveUpGitHub/)

docs/其他/Archived/AI/Machine Learning 2024-4-16.md

@@ -0,0 +1,19 @@
+# Multi-class SVM
+
+## One vs another
+
+Datasets: $C_1$, $C_2$, $C_3$
+SVMs:
+1. ($C_2$, $C_3$) vs $C_1$
+2. ($C_1$, $C_3$) vs $C_2$
+3. ($C_1$, $C_2$) vs $C_3$
+
+共 $n$ 组.
+
+## One vs one
+
+1. $C_1$ vs $C_2$
+2. $C_2$ vs $C_3$
+3. $C_3$ vs $C_1$
+
+共 $C^2_n$ 组.

docs/其他/Archived/AI/Machine Learning 2024-4-2.md

@@ -0,0 +1,11 @@
+Definition of Machine Learning(ML)
+1. Without being explicitly programmed.
+2. ML is a discipline in which computers built data-based probabilistic statistical models and use them to predict and analyze data. Also called statistical machine leaning (统计机器学习).
+
+The basic classification
+1. Supervised leaning (SL, 监督学习): SL is a ML method in which the prediction models are acquired from labeled data.
+2. Unsupervised leaning (USL, 无监督学习): USL refers to the ML problem of learning predictive models from unlabeled data.
+3. Reinforcement leaning (RL, 强化学习): RL is a ML problem in which the intelligent system learns the optional behavior strategy through continues interaction with the environment.
+4. Semi-supervised leaning (半监督学习), active leaning (主动学习).
+
+[[Support Vector Machine (SVM, 支持向量机)]]

docs/其他/Archived/AI/Support Vector Machine (SVM, 支持向量机).md

@@ -0,0 +1,46 @@
+发明者: Vapinik \[苏\]
+
+Linear support vector machine(LSVM) in the linearly separable case and Hard Margin Maximization.
+
+LSVM in the linearly separable case suppose a training dataset in a feature space, is given 
+   $$T = \{(x_1. y_1), (x_2, y_2), ..., (x_n, y_n)\}$$
+   where $x_i \in X$ = $R^n$, $y_i \in Y = \{+1, -1\}$, $i=1,2,...,N$.
+
+   Definition (linear separability of dataset)
+   Given T, if there is a hyper plane (超平面)
+   $$S: \omega \cdot X + b = 0$$
+   Than can precisely divide the positive instance points and negative instance points of dataset into both sides of the hyper plane.
+
+   i.e. For all instances $i$ with $y_i=+1$, $W \cdot x + b > 0$ and for all instances with $y_i$=-1, $\omega \cdot x + b < 0$. Then the dataset T is called a linearly separate data.
+
+---
+
+```
+^
+|   \   o
+|    \     o
+| x   \  o
+|  x   \   o
+| x  x  \
++---------------->
+```
+- `o` indicates positive instance points.
+- `x` indicates negative instance points.
+
+Find:
+1. $\omega^* \cdot x + b^* = 0$
+2. $f(x) = sign(\omega^* \cdot x + b^*)$
+
+---
+
+Function margin and Geometric margin
+
+1. Function margin: with regard to the give training dataset T and the hyperplane($W$, $b$). Define the functional margin of hyperplane on sample point $(x_i, y_i)$ as
+   $$\hat{\gamma}_i = y_i(\omega \cdot x_i + b)$$
+   $$\hat{\gamma} = \min_{i=0,...,N}\hat{\gamma}_i$$
+2. Two facts
+	1. $(\omega^*, b^*)$ 与 $(a\omega^*, ab^*)$ 相同.
+	2. Distance from $(x_0, y_0)$ to the plane $\omega_1x + \omega_2y + b = 0$:
+	   $$d = \frac{|\omega_1x + \omega_2y + b|}{\sqrt{\omega_1^2 + \omega_2^2}} = \frac{|\omega_1x + \omega_2y + b|}{||\omega||}$$
+	   Vector $\vec{x}_0$ to the hyperplane $\omega_1x + \omega_2y + b = 0$:
+	   $$d = \frac{|\omega^* \cdot x_0 + b^*|}{||\omega^*||}$$