{ "cells": [ { "cell_type": "markdown", "id": "b8a1ead2-eea5-4938-bde0-cb81b7ca5296", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "# Forecasting with Transformer\n", "\n", "## Feng Li\n", "\n", "### Guanghua School of Management\n", "### Peking University\n", "\n", "### [feng.li@gsm.pku.edu.cn](feng.li@gsm.pku.edu.cn)\n", "### Course home page: [https://feng.li/forecasting-with-ai](https://feng.li/forecasting-with-ai)" ] }, { "cell_type": "code", "execution_count": 1, "id": "136dd331-8d80-44b3-808c-b5e04cdca910", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple\n", "Requirement already satisfied: torch in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (2.7.1)\n", "Requirement already satisfied: torchvision in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (0.22.1)\n", "Requirement already satisfied: filelock in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (3.16.1)\n", "Requirement already satisfied: typing-extensions>=4.10.0 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (4.12.2)\n", "Requirement already satisfied: setuptools in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (75.6.0)\n", "Requirement already satisfied: sympy>=1.13.3 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (1.14.0)\n", "Requirement already satisfied: networkx in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (3.4.2)\n", "Requirement already satisfied: jinja2 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (3.1.4)\n", "Requirement already satisfied: fsspec in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (2024.12.0)\n", "Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.6.77 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.77)\n", "Requirement already satisfied: nvidia-cuda-runtime-cu12==12.6.77 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.77)\n", "Requirement already satisfied: nvidia-cuda-cupti-cu12==12.6.80 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.80)\n", "Requirement already satisfied: nvidia-cudnn-cu12==9.5.1.17 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (9.5.1.17)\n", "Requirement already satisfied: nvidia-cublas-cu12==12.6.4.1 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.4.1)\n", "Requirement already satisfied: nvidia-cufft-cu12==11.3.0.4 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (11.3.0.4)\n", "Requirement already satisfied: nvidia-curand-cu12==10.3.7.77 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (10.3.7.77)\n", "Requirement already satisfied: nvidia-cusolver-cu12==11.7.1.2 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (11.7.1.2)\n", "Requirement already satisfied: nvidia-cusparse-cu12==12.5.4.2 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.5.4.2)\n", "Requirement already satisfied: nvidia-cusparselt-cu12==0.6.3 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (0.6.3)\n", "Requirement already satisfied: nvidia-nccl-cu12==2.26.2 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (2.26.2)\n", "Requirement already satisfied: nvidia-nvtx-cu12==12.6.77 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.77)\n", "Requirement already satisfied: nvidia-nvjitlink-cu12==12.6.85 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (12.6.85)\n", "Requirement already satisfied: nvidia-cufile-cu12==1.11.1.6 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (1.11.1.6)\n", "Requirement already satisfied: triton==3.3.1 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torch) (3.3.1)\n", "Requirement already satisfied: numpy in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torchvision) (1.26.0)\n", "Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from torchvision) (10.2.0)\n", "Requirement already satisfied: mpmath<1.4,>=1.1.0 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from sympy>=1.13.3->torch) (1.3.0)\n", "Requirement already satisfied: MarkupSafe>=2.0 in /home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages (from jinja2->torch) (3.0.2)\n", "\n", "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m24.2\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m25.3\u001b[0m\n", "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpython -m pip install --upgrade pip\u001b[0m\n", "Note: you may need to restart the kernel to use updated packages.\n" ] } ], "source": [ "pip install torch torchvision --break-system-packages" ] }, { "cell_type": "markdown", "id": "b456061e-aadd-4a7e-8708-3bccfa1462ad", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## 生成时间序列数据\n", "- 正弦波提供周期性规律(如季节性)\n", "- 噪声代表现实中的随机波动(如市场波动)\n", "- 这相当于我们要预测的“历史销售量”或“股票价格”" ] }, { "cell_type": "code", "execution_count": 2, "id": "7ad99302-c7db-4692-8283-a78e44b4a7f9", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import torch\n", "import torch.nn as nn\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "\n", "# 1️⃣ Generate synthetic data\n", "T = 500\n", "t = np.arange(0, T)\n", "x = np.sin(0.02 * t) + 0.1 * np.random.randn(T)\n", "\n", "plt.plot(t, x)\n", "plt.title(\"Synthetic Time Series\")\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "dfb3f25f-3a83-483f-ad6b-fd6b6cacb223", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "bcb7ef78-cd0e-4479-b873-0e6c772e67a5", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "# 转换成监督学习数据\n", "- 我们把时间序列转成“特征—目标”形式。\n", "- 每一段长度为 L=20 的历史数据是输入(X);\n", "- 之后的一个点是输出(Y)。\n", "\n", "- 这就是把时间序列问题转化为机器学习能理解的监督学习任务。" ] }, { "cell_type": "code", "execution_count": 3, "id": "c222f768-d528-496c-834a-dd910060843b", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [], "source": [ "# 2️⃣ Create supervised learning dataset\n", "def create_dataset(series, L=20):\n", " X, Y = [], []\n", " for i in range(len(series) - L):\n", " X.append(series[i:i+L])\n", " Y.append(series[i+L])\n", " return np.array(X), np.array(Y)\n", "\n", "L = 20\n", "X, Y = create_dataset(x, L)\n", "X = torch.tensor(X).float().unsqueeze(-1) # (N, L, 1)\n", "Y = torch.tensor(Y).float().unsqueeze(-1) # (N, 1)" ] }, { "cell_type": "markdown", "id": "b086b28c-f7fb-4911-affd-34e5abce3dc1", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## 划分训练集与测试集\n", "- 我们用前 80% 的数据训练模型(让它“学习规律”),\n", "\n", "- 用后 20% 的数据做测试(检验它是否能“预测未来”)。" ] }, { "cell_type": "code", "execution_count": 4, "id": "0457e9eb-ad38-475f-a324-332bba54c7ea", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train size: 384, Test size: 96\n" ] } ], "source": [ "# 3️⃣ Split train/test\n", "train_size = int(0.8 * len(X))\n", "X_train, X_test = X[:train_size], X[train_size:]\n", "Y_train, Y_test = Y[:train_size], Y[train_size:]\n", "print(f\"Train size: {len(X_train)}, Test size: {len(X_test)}\")" ] }, { "cell_type": "markdown", "id": "74e31d7a-9f95-4ce8-957d-77430ab81b83", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "# 定义 Transformer 模型\n", "\n", "结构分三部分:\n", "\n", "- 输入投影 (input_proj):把标量数据(单个数)映射成“向量表示”;\n", "\n", "- Transformer Encoder:捕捉序列中不同时间点之间的依赖关系;\n", "\n", "- 解码层 (decoder):输出未来一步的预测。" ] }, { "cell_type": "code", "execution_count": 5, "id": "9fa26cc5-517e-40a3-95f7-3cb6219809d6", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [], "source": [ "# 4️⃣ Define Transformer model\n", "class TimeSeriesTransformer(nn.Module):\n", " def __init__(self, input_size=1, d_model=64, nhead=4, num_layers=2):\n", " super().__init__()\n", " self.input_proj = nn.Linear(input_size, d_model)\n", " encoder_layer = nn.TransformerEncoderLayer(\n", " d_model=d_model, nhead=nhead, dim_feedforward=128, dropout=0.1\n", " )\n", " self.encoder = nn.TransformerEncoder(encoder_layer, num_layers=num_layers)\n", " self.decoder = nn.Linear(d_model, 1)\n", "\n", " def forward(self, src):\n", " src = self.input_proj(src) # (batch, L, d_model)\n", " src = src.permute(1, 0, 2) # (L, batch, d_model)\n", " memory = self.encoder(src) # (L, batch, d_model)\n", " out = self.decoder(memory[-1]) # last token\n", " return out" ] }, { "cell_type": "markdown", "id": "7a40003c-d8f1-429e-9e52-4f4d4673b5a4", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## 参数解释\n", "\n", "### `input_size = 1`:每个时间点的输入维度\n", "\n", "* 每个时刻只有**一个特征**(比如销售量、价格、温度等)。\n", "* 如果我们有多个特征(例如价格、销量、广告支出),`input_size` 就等于特征数,比如 3 或 5。\n", "\n", "\n", "> 模型每次看到的“信息”只有一个数,就像只读“销售量”这一个指标。\n", "\n", "> 每天你只看“销售额”这一个指标在变动,没有广告、价格、天气等额外信息。" ] }, { "cell_type": "markdown", "id": "14b07ccf-60a7-45b1-a9e5-8a350358993e", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "### `d_model = 64`:内部表示的维度 (embedding dimension)\n", "\n", "* Transformer 不直接处理原始数字,而是把每个输入映射到一个 **64维向量空间**;\n", "* 这叫“特征嵌入(embedding)”,能让模型捕捉更复杂的模式;\n", "* 越高维,模型表达能力越强,但计算量也更大。\n", "\n", "🧠 **通俗理解:**\n", "\n", "> 把一个数字转成 64 维的“语义向量”,模型在这个空间里找规律。\n", "> 就像把“销售额”这个数转成一份更丰富的“销售特征报告”——里面包含周期性、波动趋势等 64 种特征维度。" ] }, { "cell_type": "markdown", "id": "eb0bad47-6481-4d37-9874-683c5f460c1a", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "### `nhead = 4`:注意力头(multi-head attention)数量\n", "\n", "* Transformer 的核心是“多头注意力机制”;\n", "* 每个 head 关注时间序列的不同方面,例如:\n", "\n", " * head 1 关注短期波动;\n", " * head 2 关注长期趋势;\n", " * head 3 关注异常跳动;\n", " * head 4 关注周期性。\n", "* 多头机制能让模型**并行地捕捉多种依赖关系**。\n", "\n", "🧠 **通俗理解:**\n", "\n", "> 模型有 4 双“眼睛”,同时看时间序列的不同特征。\n", "> 就像企业的 4 位分析师:一个看月度趋势、一个看季节波动、一个看假日效应、一个看竞争对手动态。" ] }, { "cell_type": "markdown", "id": "cc81e54c-c541-4caf-a495-afc8d6fe81db", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "### `num_layers = 2`:Transformer 编码层的层数\n", "\n", "* 每层都包含注意力机制和前馈网络;\n", "* 层数越多,模型越“深”,可以捕捉更复杂的模式;\n", "* 对于小数据(如示例),2 层就足够。\n", "\n", "🧠 **通俗理解:**\n", "\n", "> 模型有两层“思考单元”,先学到简单模式,再逐层抽象出更高层规律。\n", "> 就像公司报告流程:一层是部门分析(第一层),另一层是高管总结(第二层),信息经过两轮加工变得更有洞察力。" ] }, { "cell_type": "markdown", "id": "1f59a8b1-216a-44ef-8a85-d9bf177da16f", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## Q (Query), K (Key), V (Value) 在哪里?\n", "\n", "在我们的这段代码中(`TimeSeriesTransformer`),虽然我们没有手动写出 Q、K、V 的矩阵,但它们其实**自动在 PyTorch 的 `TransformerEncoderLayer` 内部实现**了。\n", "\n", "* 我们的 `src`(输入序列特征)会被线性投影到 Q、K、V 三个空间;\n", "* `MultiheadAttention` 会并行计算多个头(我们设了 `nhead=4`);\n", "* 每个头独立生成自己的注意力权重;\n", "* 然后所有头的结果会拼接起来,形成最终的注意力输出。" ] }, { "cell_type": "markdown", "id": "78b30e12-560f-442a-aac9-1a3a8a8ecc4d", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## 定义损失函数\n", "\n", "* 创建一个 Transformer 模型对象;\n", " * 模型包含我们之前定义的结构(输入层 → 注意力层 → 解码层);\n", " * 这个模型能从历史时间序列中“学习规律”,输出未来的预测。\n", "\n", "* 定义损失函数(loss function),即模型预测误差的衡量方式;\n", " * `MSELoss` 表示 **均方误差(Mean Squared Error)**:\n", " * 损失越小,说明预测越准确。\n", "\n", "* 定义**优化器(optimizer)**,控制模型如何学习、如何更新参数;\n", " * `Adam` 是一种自适应学习算法,能自动调整每个参数的学习步长;\n", " * `lr=1e-3` 表示学习率(learning rate)= 0.001,控制更新速度。" ] }, { "cell_type": "code", "execution_count": 6, "id": "60545ba0-54fb-44fe-ace9-0459de094498", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/fli/.virtualenvs/python3.12/lib/python3.12/site-packages/torch/nn/modules/transformer.py:382: UserWarning: enable_nested_tensor is True, but self.use_nested_tensor is False because encoder_layer.self_attn.batch_first was not True(use batch_first for better inference performance)\n", " warnings.warn(\n" ] } ], "source": [ "model = TimeSeriesTransformer()\n", "criterion = nn.MSELoss()\n", "optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)" ] }, { "cell_type": "markdown", "id": "ad7bbd6c-ceec-486a-b3d7-b32c0da17c2c", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "# 模型训练\n", "\n", "我们用训练数据反复优化模型参数,让预测误差(MSE)最小。\n", "\n", "- `loss` 衡量预测与真实值的差距;\n", "\n", "- `optimizer` 用梯度下降更新模型;\n", "\n", "- 每隔 20 轮,打印训练集与测试集误差。" ] }, { "cell_type": "code", "execution_count": 7, "id": "e6fe4fe7-e4ae-4a6c-9a59-92e3ada1c9df", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 20: train loss=0.0328, test loss=0.0415\n", "Epoch 40: train loss=0.0274, test loss=0.0225\n", "Epoch 60: train loss=0.0220, test loss=0.0249\n", "Epoch 80: train loss=0.0231, test loss=0.0236\n", "Epoch 100: train loss=0.0214, test loss=0.0230\n" ] } ], "source": [ "# 5️⃣ Training loop\n", "for epoch in range(100):\n", " model.train()\n", " optimizer.zero_grad()\n", " output = model(X_train)\n", " loss = criterion(output, Y_train)\n", " loss.backward()\n", " optimizer.step()\n", "\n", " if (epoch+1) % 20 == 0:\n", " model.eval()\n", " with torch.no_grad():\n", " val_pred = model(X_test)\n", " val_loss = criterion(val_pred, Y_test)\n", " print(f\"Epoch {epoch+1}: train loss={loss.item():.4f}, test loss={val_loss.item():.4f}\")" ] }, { "cell_type": "markdown", "id": "cd0c48f6-abdf-4fb8-b751-62c47a76206d", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "# 在测试集上预测\n", "\n", "- 我们关闭训练(no_grad)\n", "- 在没见过的数据上预测\n", "- 模型用测试集输入(历史窗口),输出对应的下一步预测值" ] }, { "cell_type": "code", "execution_count": 8, "id": "74b47cdb-c7f1-4c34-af45-d7ee273aa0dc", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [], "source": [ "# 6️⃣ One-step ahead predictions on test set\n", "model.eval()\n", "with torch.no_grad():\n", " preds_test = model(X_test).squeeze().numpy()" ] }, { "cell_type": "markdown", "id": "1f76cff4-1b29-4246-a102-bb5f6f80054a", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "source": [ "## 可视化结果\n", "\n", "- 蓝线:真实时间序列;\n", "\n", "- 红线:测试期预测;\n", "\n", "- 灰线:训练/测试分界线。" ] }, { "cell_type": "code", "execution_count": 9, "id": "79ccb249-5fb9-4a5a-b6b3-a0eabd1d46d2", "metadata": { "editable": true, "slideshow": { "slide_type": "slide" }, "tags": [] }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# 7️⃣ Plot true vs predicted on test portion\n", "plt.figure(figsize=(10,5))\n", "plt.plot(range(len(x)), x, label=\"True Series\", alpha=0.6)\n", "plt.plot(range(train_size+L, T), preds_test, label=\"Predicted (test)\", color=\"red\")\n", "plt.axvline(train_size+L, color=\"gray\", linestyle=\"--\", label=\"Train/Test split\")\n", "plt.legend()\n", "plt.title(\"Transformer One-Step Forecasting\")\n", "plt.show()" ] } ], "metadata": { "kernelspec": { "display_name": "/home/fli/.virtualenvs/python3.12", "language": "python", "name": "python3.12" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.3" } }, "nbformat": 4, "nbformat_minor": 5 }