Little-girl-1992 / RAE

Licence: other
基于tensorflow搭建的神经网络recursive autuencode,用于实现句子聚类

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to RAE

Cancerify
Turn an innocent text into torturous hell
Stars: ✭ 44 (+266.67%)
Mutual labels:  text
SHARP
SHARP: Single-cell RNA-seq Hyper-fast and Accurate processing via ensemble Random Projection
Stars: ✭ 14 (+16.67%)
Mutual labels:  clustering
RubyTextMeshPro
Unity Text Mesh Proでルビ(フリガナ)のタグを追加しました.
Stars: ✭ 61 (+408.33%)
Mutual labels:  text
ParallelKMeans.jl
Parallel & lightning fast implementation of available classic and contemporary variants of the KMeans clustering algorithm
Stars: ✭ 45 (+275%)
Mutual labels:  clustering
M-NMF
An implementation of "Community Preserving Network Embedding" (AAAI 2017)
Stars: ✭ 119 (+891.67%)
Mutual labels:  clustering
instagram-text-editor
An Instagram like text editor Flutter widget that helps you to change your text style.
Stars: ✭ 66 (+450%)
Mutual labels:  text
react-native-styled-text
Styled Text for React Native
Stars: ✭ 57 (+375%)
Mutual labels:  text
Compressed2TXT
File(s)/Folder(s) "Send to" menu .bat ascii encoder with optional password and makecab lzx compression
Stars: ✭ 156 (+1200%)
Mutual labels:  text
nbodykit
Analysis kit for large-scale structure datasets, the massively parallel way
Stars: ✭ 93 (+675%)
Mutual labels:  clustering
rake new2
A Python library that enables smooth keyword extraction from any text using the RAKE(Rapid Automatic Keyword Extraction) algorithm.
Stars: ✭ 23 (+91.67%)
Mutual labels:  text
VOSviewer-Online
VOSviewer Online is a tool for network visualization. It is a web-based version of VOSviewer, a popular tool for constructing and visualizing bibliometric networks.
Stars: ✭ 44 (+266.67%)
Mutual labels:  clustering
link text
Easy to use text widget for Flutter apps, which converts inlined urls into working, clickable links
Stars: ✭ 20 (+66.67%)
Mutual labels:  text
TrajSuite
TrajSuite is a cross-platform Java application that provides a suite of trajectory data-mining and visualisation features.
Stars: ✭ 15 (+25%)
Mutual labels:  clustering
textics
📉 JavaScript Text Statistics that counts lines, words, chars, and spaces.
Stars: ✭ 36 (+200%)
Mutual labels:  text
quartz-scheduler-hazelcast-jobstore
An implementation of a Quartz Scheduler JobStore using Hazelcast distributed Collections
Stars: ✭ 42 (+250%)
Mutual labels:  clustering
dobbi
An open-source NLP library: fast text cleaning and preprocessing
Stars: ✭ 21 (+75%)
Mutual labels:  text
mg
OpenBSD Mg editor. Portable Public Domain Micro Emacs for *BSD, Cygwin, Linux, Mac OS X.
Stars: ✭ 99 (+725%)
Mutual labels:  text
algorithms
The All ▲lgorithms documentation website.
Stars: ✭ 114 (+850%)
Mutual labels:  clustering
scarf
Toolkit for highly memory efficient analysis of single-cell RNA-Seq, scATAC-Seq and CITE-Seq data. Analyze atlas scale datasets with millions of cells on laptop.
Stars: ✭ 54 (+350%)
Mutual labels:  clustering
text-classification-baseline
Pipeline for fast building text classification TF-IDF + LogReg baselines.
Stars: ✭ 55 (+358.33%)
Mutual labels:  text

RAE(recursive autoencoder)

这个模型是将recursive NN结合Autoencoder构成一个无监督的句子转换模型。可以将不定长的句子转换成句子向量。 具体算法如下: 准备工作:将句子的词训练成词向量

  1. 将句子里面的词的词向量,两两结合,形成一个词向量对
  2. 用这些词向量对去初始化Autoencoder,生成模型和参数
  3. 将句子中邻近的词结合,输入初始化训练的Autoencoder模型,选择还原度最高的两个词生成的隐藏层输出,即x3,x4和对应的y1。
  4. 将y1代替句子中的x3,x4,重复步骤3,直到句子所有词合并完成
  5. 将选出的句子对,即[x3,x4],[x2,y1],[x1,y2]放入模型Autoencoder继续训练,更新参数
  6. 直到所有还原度都大于阈值为止
  7. 保留模型参数,计算句子向量

参考本人博客链接地址:http://blog.csdn.net/qq_26609915/article/details/52119512

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].