context position weighting

Introduction In this article, we’ll talk about challenges or limitations of word2vec model. Recall that our neural network had two weight matrices–a hidden layer and output layer. Both of these layers would have a weight matrix with 300 x 10,000 = 3 million weights each. These results in a skip-gram neural network that contains a …

Challenges in word2vec Model Read More »