Word Sequence Decoding in Seq2Seq Architectures

https://towardsdatascience.com/word-sequence-decoding-in-seq2seq-architectures-d102000344ad

1*ZiaGQhQhmN8yAK5WLeqpnA.jpeg

Word Sequence Decoding in Seq2Seq Architectures – Towards Data Science
A concise introduction to decoding algorithms in Natural Language Generation. Beam Decoding with k=3 at each time-step. Here, we search for a high scoring output sequence by keeping track of top k vocabulary outputs at each time-step while decoding.We usually stop our search when end-of-sentence token () is reached or till time-step t for all or at-least some n output sequences.
towardsdatascience.com

AutoViz: A New Tool for Automated Visualization

https://towardsdatascience.com/autoviz-a-new-tool-for-automated-visualization-ec9c1744a6ad

1*lgVdi9ZpL5_kK8s9ccPNtw.png

AutoViz: A New Tool for Automated Visualization | Dan Roth Data Science
Data scientists are often tasked with working through massive data stores to provide workable insights. These insights are then analyzed in order to identify patterns related to business intelligence or even human behavior. However, it may be one thing to construct data queries and machine learning pipelines, employing all types of optimizations and clever algorithms.
towardsdatascience.com