Tagged in

Machine Learning


fastText++: Batteries Included

In my last post, I gave an overview of how fastText was implemented, and shed some light on the design decisions made. fastText’s primary innovation is enriching the word2vec model with subword information. We need not be limited to just learning on words though. fastText…


Day 96 — WGAN for MNIST

今日主題:使用WGAN模擬MNIST資料

參考資料

  1. eriklindernoren/Keras-GAN

筆記

理論部分在 Day59 介紹過。今天來還債寫Code Study。

  • 所需的套件
from keras.datasets…

Vocab Model Specification

In this post we began exploring the effects of limiting vocabulary on a natural language task: sentiment analysis of movie reviews.

The sentiment analysis model is based on a neural architecture using two LSTM layers. Conceptually, the network reads in a…