{"id":18276350,"url":"https://github.com/yinleicoder/tensorflowjs-learn","last_synced_at":"2025-07-04T11:35:02.547Z","repository":{"id":42245445,"uuid":"283380518","full_name":"yinleiCoder/Tensorflowjs-Learn","owner":"yinleiCoder","description":"Tensorflowjs学习","archived":false,"fork":false,"pushed_at":"2022-12-30T16:53:30.000Z","size":38017,"stargazers_count":3,"open_issues_count":9,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-02-14T22:35:23.718Z","etag":null,"topics":["tensorflowjs"],"latest_commit_sha":null,"homepage":"","language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/yinleiCoder.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2020-07-29T02:43:02.000Z","updated_at":"2023-11-27T13:45:45.000Z","dependencies_parsed_at":"2023-01-31T12:00:22.751Z","dependency_job_id":null,"html_url":"https://github.com/yinleiCoder/Tensorflowjs-Learn","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yinleiCoder%2FTensorflowjs-Learn","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yinleiCoder%2FTensorflowjs-Learn/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yinleiCoder%2FTensorflowjs-Learn/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/yinleiCoder%2FTensorflowjs-Learn/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/yinleiCoder","download_url":"https://codeload.github.com/yinleiCoder/Tensorflowjs-Learn/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247974734,"owners_count":21026742,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["tensorflowjs"],"created_at":"2024-11-05T12:15:49.598Z","updated_at":"2025-04-09T04:18:35.597Z","avatar_url":"https://github.com/yinleiCoder.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Tensorflow.js机器学习\n\n### 1. Tensor是什么\n\n### 2. 线性回归\n\n- 定义模型结构\n- 损失函数\n- 优化器：随机梯度下降\n- 训练模型\n- 进行预测\n\n### 3. 归一化\n\n- 归一化任务\n- 归一化训练数据\n- 训练、预测、反归一化\n\n### 4.逻辑回归：输出的是概率\n\n- 逻辑回归任务\n- 加载二分数据\n- 定义模型结构：带有激活的数的单个神经元\n- 损失函数：对数损失log loss \n- 训练模型并可视化训练过程\n- 进行预测\n- 二分类数据集生成函数\n\n### 5. 多层神经网络\n\n- 多层神经网络任务\n- 加载XOR数据集（XOR异或逻辑回归）\n- 定义数据模型：多层神经网络\n- 训练模型并预测\n\n### 6.多分类\n\n- 多分类任务\n- 加载iris数据集(训练集和验证集)\n- 定义模型结构：带有softmax的多层神经网络\n- 训练模型：交叉熵损失函数与准确度度量\n- 定义分类预测方法\n\n### 7.欠拟合与过拟合\n\n- 欠拟合与过拟合任务\n- 加载带有噪音的二分类数据\n- 简单神经网络演示欠拟合\n- 复杂神经网络演示过拟合\n-  过拟合应对法：早停法、权重衰减、丢弃法\n\n### 8.卷积神经网络CNN识别手写数字： 卷积层(提取特征)、池化层(提取最强的特征)、全连接层(dense作为输出层、作为分类器)\n\n- 使用卷积神经网络识别手写数字任务\n- 加载MNIST数据集\n- 定义模型结构：卷积神经网络\n- 训练模型\n- 进行预测\n\n### 9.预训练模型进行图片分类\n\n- 预训练模型进行图片分类任务\n- 加载MobileNet模型（MobileNet是一种卷积神经网络）\n- 进行预测\n\n\u003e 预训练模型：\n\u003e\n\u003e 已经事先训练好的模型，无需训练即可预测\n\u003e\n\u003e tensorflow.js中可以调用web格式的模型文件\n\n### 10.迁移学习的图片分类器\n\n- 基于迁移学习的图像分类器：商标识别任务\n- 加载商标训练数据并可视化\n- 定义模型结构：截断模型+双层神经网络\n- 迁移学习下的模型训练\n- 迁移学习下的模型预测\n\n\u003e 迁移学习：\n\u003e\n\u003e 把已训练好的模型参数迁移到新的模型来帮助新模型的训练\n\u003e\n\u003e 深度学习模型参数多，从头训练成本高\n\u003e\n\u003e 删除原始模型的最后一层，基于此截断模型的输出训练一个新的（通常相当浅的）模型\n\n### 11.基于预训练模型或迁移学习进行语音识别\n\n加载预训练语音识别模型(Tensorflow官方)\n\n使用tfjs-models的speech-commands包加载模型\n\n迁移学习：声控轮播图\n\n### 12. python模型与js模型互转：\n\n- python模型：通过python版Tensorflow/keras生成模型\n\n- python模型包括：Tensorflow SavedModel、Keras HDF5 model等\n- 获取：开源网站下载/算法同事提供\n- js模型包括：tfjs_layers_model、tfjs_graph_model等\n- 获取：通过开源网站/通过TFJS生成/由Python模型转化而来\n\n怎么互转：\n\n1. 安装tensorflow.js Converter\n2. 在命令行指定输入输出的路径和模型格式即可。\n\npython与js互转、js模型的互转：分片、量化、加速","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyinleicoder%2Ftensorflowjs-learn","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fyinleicoder%2Ftensorflowjs-learn","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyinleicoder%2Ftensorflowjs-learn/lists"}