MuhammedBuyukkinaci/TensorFlow-Sentiment-Analysis-on-Amazon-Reviews-Data
Implementing different RNN models (LSTM,GRU) & Convolution models (Conv1D, Conv2D) on a subset of Amazon Reviews data with TensorFlow on Python 3. A sentiment analysis project.
Leverages pre-trained GloVe embeddings (100D) and compares 13 distinct model variants, including bidirectional architectures and attention mechanisms, with hardware-specific optimizations (CUDNN for GPU, CPU-tuned implementations). Features custom early stopping logic that monitors test loss trends across 80-epoch windows, and combines sequential layers with Conv1D/Conv2D feature extraction for binary sentiment classification on 150k training samples.
No commits in the last 6 months.
Stars
54
Forks
18
Language
Python
License
—
Category
Last pushed
Sep 21, 2019
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/MuhammedBuyukkinaci/TensorFlow-Sentiment-Analysis-on-Amazon-Reviews-Data"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
3inchtime/douban_sentiment_analysis
基于朴素贝叶斯实现的豆瓣影评情感分析
Robin-WZQ/Text-sentiment-polarity-judgment
基于规则、基于朴素贝叶斯、基于逻辑回归进行文本情感极性分析判断（酒店评论语料）
nhtlongcs/StarListify
StarListify is a Python package that classifies GitHub stars history into organized category...
sharmaroshan/Employee-Reviews
This is Project which contains Data Visualization, EDA, Machine Learning Modelling for Checking...
sharmaroshan/Coursera-Reviews-Analysis
It is a Natural Language Processing Problem where we have to decide the sentiments of the users...