《python深度学习》由Keras之父、现任Google人工智能研究员的弗朗索瓦•肖莱(François Chollet)执笔,详尽介绍了用Python和Keras进行深度学习的探索实践,包括计算机视觉、自然语言处理、生成式模型等应用。书中包含30多个代码示例,步骤讲解详细透彻。
以下代码包含了全书80%左右的知识点,代码目录:
- 2.1: A first look at a neural network( 初识神经网络)
- 3.5: Classifying movie reviews(电影评论分类:二分类问题)
- 3.6: Classifying newswires(新闻分类:多分类问题 )
- 3.7: Predicting house prices(预测房价:回归问题)
- 4.4: Underfitting and overfitting( 过拟合与欠拟合)
- 5.1: Introduction to convnets(卷积神经网络简介)
- 5.2: Using convnets with small datasets(在小型数据集上从头开始训练一个卷积
- 5.3: Using a pre-trained convnet(使用预训练的卷积神经网络)
- 5.4: Visualizing what convnets learn(卷积神经网络的可视化)
- 6.1: One-hot encoding of words or characters(单词和字符的 one-hot 编码)
- 6.1: Using word embeddings(使用词嵌入)
- 6.2: Understanding RNNs(理解循环神经网络)
- 6.3: Advanced usage of RNNs(循环神经网络的高级用法)
- 6.4: Sequence processing with convnets(用卷积神经网络处理序列)
- 8.1: Text generation with LSTM(使用 LSTM 生成文本)
- 8.2: Deep dream(DeepDream)
- 8.3: Neural style transfer( 神经风格迁移)
- 8.4: Generating images with VAEs(用变分自编码器生成图像)
- 8.5: Introduction to GANs(生成式对抗网络简介)
作者的github:https://github.com/fchollet/deep-learning-with-python-notebooks