I am trying out a longer term (3-4 month) planning methodology which might affect the way I look at weekly reviews, specially Experiments section which stays quiet on a shorter scale. Will have some thoughts next week.
1 Experiments
2 Readings/Explorations
- Mixing dirichlet topic models and word embeddings to make lda2vec (moody2016mixing).
- Performance matters.
- A little bit on user stories.
- A really exciting Quantum Supremacy update.
3 Programming
Commits for week 38-2019 and 4 previous weeks.